Astm stp 369 1965

VACUUM ANNEALING Materials and Procedures: The experimental materials comprised cold rolled and annealed 0.020-in.-thick by 3-in.-wide stock from three different heats of AISI 301 stai

ADVANCES IN THE TECHNOLOGY

OF STAINLESS STEELS AND

RELATED ALLOYS

A compilation of papers presented at four symposiums:

"Advances in the Technology of Stainless Steels and Related Alloys," sponsored

by the American Society for Testing and Materials, Atlantic City, N J , June, 1963;

"Recent Advances in the Metallurgy of Stainless Steels," sponsored by the

Metal-lurgical Society of the American Institute of Mining, MetalMetal-lurgical, and

Petro-leum Engineers, Cleveland, Ohio, October, 1963; and "Stress Corrosion Cracking"

and "Evaluation Tests for Stainless Steels," sponsored by the National

Associa-tion of Corrosion Engineers in collaboraAssocia-tion with The Electrochemical Society,

New York, N.Y., March, 1963

Reg U.S Pat Off

ASTM Special Technical Publication No j6g

Price $21.50; to ASTM, NACE, AIME, and ECS members Si5.00

Published by the AMERICAN SOCIETY FOR TESTING AND MATERIALS

1916 Race St., Philadelphia 3, Pa

© BY AMERICAN SOCIETY FOR TESTING AND AIATERIALS 1965

Library of Congress Catalog Card Xumber: 65-19685

Printed in Baltimore, Md

April, 1965

F O R E W O R D

During 1963 the American Society for Testing and Materials; the Metallurgical Society of the American Institute of Mining, Metallurgical, and Petroleum Engineers; and The Electrochemical Society and the National Association of Corrosion Engineers, organized symposia dealing with several aspects of the technology of stainless steels Members of these organizations who were concerned with these activities decided that

it would be advantageous to those interested in stainless steels if the papers presented

at these meetings could be combined in a single pubhcation for convenient future ence They formed an Intersociety Co-ordinating Committee to accomphsh this

refer-With the collaboration of the authors of the several papers, the participating societies, and particularly the American Societ\- for Testing and Materials which undertook publi-cation of the papers in this volume, the desired consolidation of papers from the several societies has been achieved

The symposia in which the several papers originated were as follows: (1) symposium

on Advances in the Technology of Stainless Steels and Related Alloys, Atlantic City, N J.,

June, 1963, Sponsored by ASTM Committee A-10, M A Cordovi, chairman; (2)

sym-posium on Stress Corrosion Cracking, H L Logan, E H Phelps, and H R Copson, co-chairmen, and symposium on Evaluation Tests for Stainless Steels, R B Mears,

chairman Second International Congress on Metallic Corrosion Sponsored by NACE

in collaboration with The Electrochemical Soc, New York, N Y., March, 1963; and

(3) symposium on Recent Advances in the Metallurgy of Stainless Steels, Cleveland, Ohio,

October, 1963, sponsored by the Metallurgical Society of AIME, E J Dulis and L R Scharfstein, co-chairmen

The task force concerned with this activity wishes to express its thanks to all who have co-operated in this venture It is hoped that the results of this effort will be well received by those in whose interests it was undertaken and that its success will prompt similar future action in consoHdating information from several sources in single publica-tions for more convenient reference

The Intersociety Co-ordinating Committee consisted of M A Cordovi, ASTM;

E J Duhs, AIME; R B Mears, NACE; and E H Phelps, ECS, with F L LaQue, chairman

NOTE—The Society is not responsible, as a body, for the statements

and opinions advanced in this publication

CONTENTS

Advances In the Technology of Stainless Steels and Related Alloys

PAGE

Effects of Vacuum Melting and Vacuum Annealing on tlie Properties of Austenitic Stainless Steels—

D B Roach, W B Lefiingwell, and A M Hall 1 Effects of Austrolling on the Properties of Crucible 422 Stainless Steel—R C VVestgren and E J

Dulis 8 Discussion 16 High-Strength Stainless Steel by Deformation and Low Temperatures—S Floreen and J R

Mihalisin 17 The Mechanisms of Deformation and Work Hardening in AISI Type 301 Stainless Steel—W F

Barclay 26 Practical Aspects of Production Bright Annealing of Stainless Steel—R C Boyer and R J Perrine 30

AdvancementsinExtrusion and Cold Finishing of Stainless Steel and Irregular Shapes—J J Barrett 36

New Martensitic Age Hardening Stainless Steels—G N Aggen, C M Hammond, and R A Lula 40

Discussion 46 The Development of New High-Strength Stainless Steels—C M Hammond 47

Mechanical Properties and Corrosion Resistance of a High-Strength

Chromium-Manganese-Nitro-gen Stainless Steel—J J Heger 54 Discussion 61 Metallurgy of a Columbium-Hardened Nickel-Chromium-Iron Alloy—H L Eiselstein 62

Discussion 78 Properties of 12 Per Cent Chromium Alloys Modified with Small Columbium Additions—H

Tanczyn 80 Discussion 87

A High-Strength Weldable Stainless Steel for Elevated-Temperature Service—F C Hull 88

Discussion 98 Creep-Rupture Properties of Stainless Steels at 1600, 1800, and 2000 F—K G Brickner, G A

Ratz, and R F Domagal 99 The Effects of Neutron Exposure and Reactor Environments on Stainless Steels—S H Bush and

J C Tobin 112 Evaluation and Application of Stainless Steels in Cryogenic Environments—A Hurlich and W G

Scheck 127 Experience with Stainless Steels in Utility Power Plants—G E Lien 136

Discussion 142 Modified Type 316 Stainless Steel with Low Tendency to Form Sigma—C E Spaeder and K G

Brickner 143 Effect of Composition and Section Size on Mechanical Properties of Some Precipitation Hardening

Stainless Steels—W C Clarke, Jr and H VV Garvin 151 New Cast High-Strength Alloy Grades by Structure Control—F H Beck, E A Schoefer, J W

Flowers, and M G Fontana 159 The Effect of Small Columbium Additions to Type 304 Stainless Steel—R B G Yeo and T E

Scott 175 Discussion 179

Stress Corrosion and Evaluation Tests for Stainless Steels

Effect of Heat Treatment and Welding on Corrosion Resistance of Austenitic Stainless Steels—J

Robert Auld 183

An Appraisal of Evaluation Tests for Stainless Steel Automotive Trim—E H Phelps and J F

Bates 200 Accelerated Tests as a Method of Predicting Service Corrosion of Exterior Automotive Trim—G F

Bush, W J Garwood, B E Tiffany 209

A Laboratory Test for Determining Susceptibility of Nickel-Molybdenum and

Nickel-Molybdenum-Chromium Alloys to Intercrystalline Corrosion and Its Use in the Development of Resistant Alloys—H Grafen 223

An Evaluation of Accelerated Strauss Testing—L R Scharfstein and C M Eisenbrown 235

The Relationship of Heat Treatment and Microstructure to Corrosion Resistance in Ni-Cr-Mo

Alloys—M A Streicher 240

A Testing System for Detecting Susceptibility to Rapid Intergranular Attack in Various Grades of

Austenitic Stainless Steels—M A Streicher 255 Effect of Heat Treatment, Composition, and Microstructure on Corrosion of 18 Cr-8 Ni-Ti Stainless

Steels in Acids—M A Streicher 257

Recent Advances In the Metallurgy of Stainless Steels

Time-Temperature-Sensitization (TTS) Diagrams for Types 347, 304L, and 316L Stainless Steels—

H F Ebling and M A Scheil 275 Discussion 283

vi COXTENTS

PAGE

Alloying Precipitation Hardening Stainless Steels for Strength and Stability—D C Perry and M W

Marshall 285 Discussion 290 Fine Structure and Properties of a 12-Cr MoVVv Martensitic Stainless Steel—B R Banerjee, J J

Hauser, and J M Capenos 291 Further Studies on the Formation of Sigma in 12 to 16 Per Cent Chromium Steels—D C Ludwig-

son and H S Link 259 Discussion 311 Development of a Modified Alloy 20 Stainless Steel—H L Black and L W Lherbier 312

Discussion 318 Relationship Between Metallurgical Structure and Properties of a Precipitation Hardening Stain-

less Steel—G N Aggen and R H Kaltenhauser 319 Aging Mechanisms in Precipitation-Hardening Stainless Steels—J C Wilkins, R E Pence, and

D C Perry 331 Discussion 341

A Study of Internal-Friction Peaks in T>'pe 304 Stainless Steel Containing Nitrogen—J F Eckel and C R Manning, Jr 342 Discussion 347 Roping and Intergranular Corrosion of 430 Stainless Steel—E A Parker 348

RELATED ASTM PUBLICATIONS

Stress-Corrosion Cracking of Austenitic Chromium-Nickel Stainless Steels, STP 264 (1960) Comparison of the Properties of Basic Oxjgen and Open Hearth Steels, STP 364 (1963) Trends in the Metallurg\- of Low-Alloy, High-Yield-Strength Steels, Gillette Lecture (1963)

Advances in the Technology

of Stainless Steels and Related

Alloys

STP369-EB/Apr 1965

E F F E C T S OF VACUUM M E L T I N G AND VACUUM A N N E A L I N G

ON T H E P R O P E R T I E S OF A U S T E N I T I C STAINLESS STEELS

B Y D B ROACH;! W B LEFFINGWELLI^ AND A M HALL,^

Personal Member, ASTM SYNOPSIS

Two applications of vacuum technology to metallurgy are discussed One

concerns the consumable-electrode vacuum-arc remelting of AISI 316 stainless

steel An air-melted and a vacuum-arc remelted heat are compared with

re-spect to composition, processing and welding, room- and high-temperature

mechanical properties, magnetic permeability, inclusion count, and behavior

in the Huey corrosion test The vacuum heat was lower in carbon, oxygen,

and hydrogen than the air-melted heat The low oxygen and h3'drogen were

attributed to the melting process, while the low carbon was not Both heats

were similar in processing and welding behavior as well as in mechanical and

physical properties In Huey corrosion tests of sensitized material, the

air-melted steel was attacked much faster than the vacuum-air-melted material

However, this difference was ascribed to the difference in carbon content

between the two steels Thus, in this study, no differences attributable to the

consumable-electrode vacuum-arc remelting process were observed

The other application was the vacuum annealing of AISI 301 strip This

material was compared with air-annealed and hydrogen-annealed stock The

air-annealed and vacuum-annealed steels had similar mechanical properties

However, the hydrogen-annealed steel had half of the ductility, 80 per cent

of the formability and 75 per cent of the strength of the other materials The

increase in hydrogen content of the steel, when annealed in hydrogen, is

con-sidered responsible for these differences

Vacuum technology has been p u t to

work at a number of points in the

con-solidation, refining, working, fabrication,

and heat treatment of a considerable

variety of metals and alloys The use of a

vacuum in melting operations has come

in for a particularly large share of

atten-tion Such processes as vacuum-induction

melting and consumable-electrode

vac-uum-arc remelting have been developed

and applied to numerous metallic

mate-rials The various vacuum melting

proc-esses have found principal application

in the consolidation and refining of the

reactive metals, the refractory metals,

numerous superalloys, certain specialty

steels, and a variety of ultrahigh-strength

steels Since World War I I a copious

literature has come into being on these

subjects

On the other hand, little information

is to be found in the literature regarding

the influence of vacuum melting on the

properties of the stainless steels Some

data have been reported which showed

' Research associate, Battelle Memorial Inst.,

Columbus, Ohio

^Assistant chief metallurgist, Sharon Steel

Corp., Sharon, Pa

'Division chief, Battelle Memorial Inst.,

Columbus, Ohio

Copyright® 1965 by ASTM Intemational

t h a t the ductile-to-brittle transition temperatures of vacuum - induction- melted AISI Types 430 and 446 stainless steels were well below room temperature, while the corresponding air-melted steels went through the ductile-to-brittle tran- sition considerably above room tempera- ture (l).-* Data on AISI Types 403, 410,

422, and 446 steels show that the temperature impact strength, as meas- ured by the Izod or the Charpy test, was markedly greater for vacuum-induction- melted alloys than for their air-melted counterparts (2) In other research, some

room-of the properties room-of air-melted and uum-induction-melted chromium-nickel and chromium-nickel-molybdenum stain- less steels were compared at room tem- perature and a t 1300 F (3) The data indicate that the vacuum-melted mate- rial had slightly greater room-tempera- ture tensile ductility and considerably greater rupture strength at 1300 F The steels were of the 16 per cent chromium,

vac-16 per cent nickel type containing 0.04

to 0.08 per cent carbon and 0.51 to 0.87 per cent columbium I n addition, com- parative data have been reported on

* The boldface numbers in parentheses refer

to the list of references appended to this paper

1 www.astm.org

air-melted and on consumable-electrode vacuum-arc-remelted AM 355 (4) These data indicate that the vacuum-melted material was somewhat stronger and considerably more ductile than the corre- sponding air-melted material processed according to the same schedule The gains in properties were attributed to the homogenizing action which may occur

in consumable-electrode vacuum-arc melting

re-One of the programs described in this paper was directed at casting additional light on the influence of vacuum melting

on the properties of austenitic stainless steel Specifically, material from a com- mercial heat of consumable-electrode, vacuum-arc remelted AISI 316 was com- pared with material from a commercial heat of air-melted AISI 316 The criteria for the comparison were composition, processing behavior, corrosion resistance, microstructure, mechanical properties, and weldability

The other program discussed here was

a brief study of the influence on the tensile properties and cold-forming capa- bihty of chromium-nickel stainless steel strip produced by the environment used

in annealing Vacuum anneahng was emphasized, and the material under study was commercially produced AISI

301

CONSUMABLE-ELECTRODE

VACUUM-ARC R E M E L T I N G

Materials and Procedures:

The two heats used in this tion were made in the regular production facihties of the Sharon Steel Corp The air-melted heat was produced in a direct- arc furnace, and the vacuum-melted steel was made by the consumable- electrode method, the electrodes having been made from a standard arc-melted heat The heat analyses are given in Table 1 Except for carbon content, the differences in the heat analyses of the two steels were not considered particu- larly significant

investiga-A 9- by 2- by 60-in hot-rolled slab

ADVANCES IN THE TECHNOLOGY OF STAINLESS STEELS

from each heat was furnished as starting

material for the experimental work

Transverse sections of the slabs, 2 by 2

by 9 in., were heated at 2150 F and

forged to |-in square bars which, in

turn, were reheated to the same

tem-perature and rolled to f-in round bars

to provide material for freezing-cycle

hot tension tests Sections 22 in long

were cut from the slabs, heated to 2150 F,

and rolled to |-in plate Part of the

plate was reserved for impact and

mag-netic permeability tests, and the

re-mainder was reheated to 2000 F and

rolled to 0.10-in.-thick strip This strip

was annealed 15 min at 1950 F, water

quenched, and pickled in hot 5 per cent

hydroiluoric acid-20 per cent nitric acid

Part of it was saved for weldability tests;

part was cold rolled to 0.060-in

thick-and hydrogen contents thick-and about the same nitrogen content as the air-melted material, but noticeably higher oxygen than the corresponding 2-in.-thick mate-rial

The differences in the gas contents of the material in the two different thick-nesses appeared to be caused either by pickup or loss of the various gases during processing The air-melted steels ap-peared to lose hydrogen but to pick up nitrogen during hot working and anneal-ing The vacuum-melted steel appeared

to pick up both oxygen and nitrogen

Tensile properties were obtained for 0.060-in.- and 0.045-in.-thick material using standard sheet tension specimens

A strain rate of 0.02 min~' was used to the yield point, after which a crosshead speed of 0.2 in per min was employed

TABLE 1—HEAT ANALYSES OF T H E EXPERIMENTAL MATERIALS

" Before vacuum melting

' After vacuum melting

0 5 290

0 5 320

ness, degreased, annealed and pickled,

and machined into tensile, creep and

stress-rupture specimens; part was cold

rolled to 0.060-in thickness, degreased,

annealed, pickled and reroUed to

0.045-in thickness to provide half-hard stock

for tension tests; finally, some of the

0.045-in strip was annealed and pickled

for tension tests and corrosion tests

The experimental materials were

analyzed for oxygen, hydrogen, and

nitrogen when in the form of 2-in.-thick

slabs and 0.045-in.-thick strip The

re-sults are shown in Table 2 In the form

of 2-in.-thick slabs the vacuum melt

showed much lower oxygen and hydrogen

than the air melt and about the same

nitrogen content The 0.045-in.-thick

material from the air melt was noticeably

lower in hydrogen and somewhat higher

in nitrogen than the corresponding

2-in.-thick stock The vacuum-melted

0.045-in.-thick material showed lower oxygen

Charpy impact tests were made with the keyhole type of specimen The material investigated was annealed The holes for the notches were drilled perpendicular

to the surface of the plates

^Magnetic permeability tests were made on three specimens $ in in diameter

by 1 in in length cut from different areas

of |-in.-thick annealed plate of each of the two steels The permeability meas-urements were made with a direct-read-ing permeameter Inclusion counts were made on sections of the 2-in.-thick slab and the 0.045-in strip from the air melt and the 0.045-in.-thick strip from the vacuum melt ASTM Recommended Practice E 45, Method A, for inclusion ratings was used in this work."

The corrosion evaluation consisted

in determining the performance of the materials in the Huey 65 per cent boiling nitric acid test, as carried out according

to ASTM Recommended Practice A 262.6 Briefly, the test consisted of completely immersing triplicate 2- by j - by 0.045-in

specimens in the boiling acid solution

A 1000-ml Erlenmeyer flask and denser apparatus, heated by an elec-

con-^ Recommended Practice for Determining

the Inclusion Content of Steel, 1964 Book of ASTM Standards, Fart 31

* Recommended Practice for Boiling Nitric

Acid Test for Corrosion-Resisting Steels, 1964 Book of ASTAf Standards, Part 3

trie hot plate, were used for each grouj

of three specimens A comparable sur face finish for each specimen was ob tained by hand polishing with 120-gri abrasive paper The test consisted of fivi periods each of 48-hr duration Weigh loss was determined after each tes period

The rate of corrosion was calculatee

from R = KW/AST

where:

R = rate of corrosion in mils penetra

tion per month,

K = 43,900,

Tf = weight loss,

A — total surface area, sq in.,

S = density of the material, g pe

cu cm, and

T = duration of the test period, hr

Specimens were tested as annealec

at 1950 F for 15 min and air cooled an(

as sensitized by heating at 1200 F fo

1 hr followed by furnace cooling Three rupture tests and three creej tests were made at 1200 F on each of thi materials in the form of 0.060-in.-thicl strip Creep as well as stress-ruptun data were obtained from the ruptun tests, but only creep data were obtainec from the creep tests Stresses for thi rupture tests were chosen to produo failure in less than 1000 hr, and al creep tests were discontinued afte: about 1000 hr

Welding quality was investigated b} preparing welded joints in annealec 0.10-in.-thick strip produced from th( two heats of steel and testing them fo: bend ductility, tensile strength, anc elongation In addition, to determine i

a difference in crack sensitivity existec between weld deposits made in thesi steels, the high-temperature ductilitj

of bar stock was obtained over the tern perature range from 2550 to 2000 F the range in which cracking generall} occurs These data were obtained durinj

a freezing and cooling cycle designed t( simulate the conditions that exist durinj the cooling of a weld deposit

In studying the bend ductility o: welded joints, the guided bend tesi was used with two types of joint anc two types of specimen per joint Ir one type of specimen the weld wa; transverse to the rolling direction, anc

in the other it was parallel to the rolling direction In one type of joint no fillei

wire was used, while in the other i

small amount of filler wire was used tc insure a cross section in the weld meta

at least equal to that of the base ma terial A commercial Type 316 fiUei

ROACH ET AL ON VACUUM MELTING AND ANNEALING

wire was deposited with the

tungsten-arc process for this purpose The weld

reinforcement was left on all of the

bend specimens Tests were made with

the root of the weld in tension and also

if no failure occurred, they were

de-A-ljuslmpnt plote to allow for contfoction rjuring cooling

Water-cooled copper blocks (2)

Induction coil

Ceromic thermocouple protection tube

Extension arm connected

to load cell

FIG 1—Cross Section of Test Specimen Showing Method of Attachment to Testing Machine

TABLE 3—AVERAGE TENSILE PROPERTIES OF 0.060- AND 0.045-GAGE STRIP."

rp , Tensile Yield Elongation

Heat No imi-Kness, Condition Direction Strength, Strength, in 2 in.,

1000 psi 1000 psi percent 1-E 0.060 annealed* longitudinal 78.0 29.5 61.5

5-V 0.060 annealed longitudinal 79.0 30.5 59.5

1-E 0.045 annealed' longitudinal 80.5 31.0 62,5

5-V 0.045 annealed longitudinal 80.0 31.0 63.5

1-E 0.045 half hard' longitudinal 121 110 16.5

5-V 0.045 half hard longitudinal 120 111 18.5

1-E 0.045 annealed transverse 79.0 29.5 59.0

5-V 0.045 annealed transverse 79.5 31.0 61.5

1-E 0.045 half hard transverse 123 104 14.5

S-V 0.045 half hard transverse 118 101 16.5

° Averaged from quadruplicate tests

' 1950 F for 15 min and air cooled

' 25 per cent cold reduction

with the face of the weld in tension In formed over successively smaller dies

the longitudinal bend specimen, the until they failed The bend ductility of

weld metal, heat-affected zones, and the welded joints was calculated from

base metal were deformed together, the smallest radius over which they

while the ductility of the weld metal could be bent 180 deg before failure

only was determined from the transverse occurred

The freezing cycle tension test was developed at Battelle during the course

of a study to determine the cause of weld metal cracking in low and high alloy steel (5) The test specimen and the method of heating and loading it are shown in Fig 1 The apparatus was designed so that the strength and duc-tility of the material could be obtained over the temperature range from the solidus to room temperature Air pres-sure was used to activate an air cylinder which was connected by an extension arm and holder to the specimen being tested The load appHed to the test specimen was recorded on a high-speed recorder from strain gages attached to

a load cell connected to the lower tension arm A 30-kw spark-gap induc-tion unit was used to heat the specimens

ex-to the liquid range The specimens were heated to a temperature of 2900 F in approximately 20 sec They were then allowed to cool to various predetermined temperatures and rapidly loaded to failure on reaching the desired tempera-ture The temperature of the specimens was measured by placing a platinum-platinum-10 per cent rhodium thermo-couple in the molten zone as shown in Fig 1 The couple was protected from the molten alloy by a thin-walled ceramic tube A quartz sleeve was used to retain the molten metal while above the solidus temperature After a few tests were made and cooling curves were obtained for each, it was apparent that the heating and cooHng cycles were the same for both materials As a result, subsequent heating and coohng times were selected from these curves rather than from thermocouples placed in each specimen

To maintain a low temperature gradient across the diameter of the specimen, water-cooled copper blocks were clamped adjacent to the melt zone

on each end of the specimen By cording the temperature at the center and the outside edge of the specimen,

re-it was found that the temperature gradient across the specimen during cooling never exceeded 25 F

From the results of the freezing cycle tension test, it has been possible to determine the effects of relatively small changes in chemical composition on cracking susceptibility If the material exhibited low ductihty at elevated temperatures, the cracking resistance also was poor

Results:

Forging and Rolling—Although no

quantitative measure of forgeability

was used, it can be said that the bars

forged well and no difference in

forge-ability between tfie two heats was

ob-served Likewise, the materials

nego-tiated hot and cold rolhng successfully;

the surfaces of stock rolled from each

heat were good and no edge cracking

Average Charpy Values,'' ft-lb

ature -lOOF

terials, as annealed and as sensitized Each value shown in the table is the average of the results obtained from three specimens; the individual results were in good agreement with each other

T A B L E 7 — R U P T U R E A N D C R E E P D A T A F O R 0 0 6 0 - G A G E S T R I P A T 1200 F

Heat No Specimen No Stress, psi Rupture Time, hr

1-E 1 35 000 5 3 1-E 2 27 500 6 5 6 1-E 3 20 000 8 6 1 7 1-E 4 13 000 1 0 2 8 0 "

1-E 5 9 000 1 0 0 3 5 "

1-E 6 6 000 1 0 0 6 6 "

5-V 2 35 0 0 0 2 7 5-V 1 27 500 5 9 0 5-V 3 20 000 5 0 9 4 5-V 4 13 000 1 0 1 3 0 "

5-V 6 9 000 1 1 7 0 6 "

5-V 7 6 000 1 0 0 5 9 "

" T e s t d i s c o n t i n u e d at t h i s t i m e ' T o t a l s t r a i n a t t h e t i m e t h e creep test was d i s c o n t i n u e d

Total Elongation, per cent

Minimum Creep Rate, per cent per hour 33.7

38.9 42.1 2.50*

0.414'' 0.108'' 38.8 35.0 30.4 4.01'' 0.48'' COBS'-

2.3 0.22 0.020 0.0015 0.00026 0.00006 5.0 0.16 0.026 0.0032 0.0002 0.000035

t h i n A-2 t h i n -|- D-2 t h i n

Tensile and Impact Properties—From

Table 3 it is seen that both heats had

essentially the same tensile properties

Likewise, both had high impact

re-sistance at all test temperatures,

in-cluding — 320 F, as shown in Table 4

This behavior is characteristic of AISI

316

Magnetic Permeability and Inclusion

Count—The results of the magnetic

permeability tests showed that all of the

specimens had the same permeability of

LOOS, which indicated that all were

_

- - X " D '

0 0 0 0 0 1 OOOOt OOOI 001 01

Minimum Creep Rote , per cent per hour

F I G 3—Stress Versus M i n i m u m C r e e p R a t e for 0.060-Gage S t r i p a t 1200 F

from the inclusion counts are hsted in In addition, each value in the table is Table 5 the average of the results from five 48-hr

Performance in the Huey Ccrrosion immersion periods As annealed, the Test—Table 6 gives the average cor- two materials behaved alike However,

rosion rates obtained for the two ma- when sensitized, the vacuum-melted

ROACH ET AL ON VACUUM MELTING AND ANNEALING

steel displayed nearly ten times the

resistance of the air-melted material

This difference is attributed to the extra

low carbon content of the former

Creep and Stress-Rupture Data—The

rupture and creep data obtained are

listed in Table 7 Rupture time versus

stress is plotted in Fig 2, while

mini-mum creep rate is shown in Fig 3 as

a function of stress In each figure,

a radius less than the thickness of the sheet material without failing This amount of deformation represents ap-proximately 33 per cent elongation in the outer fibers of the bend specimen

The data indicated that there was no difference in the bend ductility of welded joints made in the two materials

Duplicate transverse and longitudinal tension specimens were machined from

t r a n s v e r s e

t r a n s v e r s e parallel parallel

b a s e m e t a l base m e t a l fusion zone base m e t a l

T A B L E 9 — T E N S I O N T E S T D A T A F O R 0 1 0 0 - G A G E S H E E T S P E C T A I E N S W E L D E D

W I T H T Y P E 316 S T A I N L E S S S T E E L F I L L E R W I R E

Heat No Specimen No

Relationship of Weld Elongation

to Direction of Rolling in 2 in.,

of Base Plate per cent

Tensile Strength, psi Location of Fracture 1-E

a single curve has been used to represent

the behavior of both heats It is

con-sidered that deviations of the individual

data points from the curves are well

within the normally expected scatter

band The ductility at rupture of the

two materials was strikingly similar for

equivalent stresses and times

Weldability Results—Sixteen bend

specimens were prepared from each of

the steels All of the different types of

bend specimens could be formed around

samples of each heat in which the welds were made without filler wire The weld reinforcement was removed The results

of the tension tests are listed in Table 8

The joint efficiency of the specimens was good In most cases, the tensile strength was equal to that of the unwelded base material In all but one of the specimens, failure occurred in the base metal

Tension test data also were obtained

on transverse welded joints prepared with Type 316 filler wire The test re-

sults are shown in Table 9 Again, the weld-joint properties were about equal

to those of the base material, and no appreciable difference was found in the mechanical properties of the two different heats of steel

The data obtained from the freezing cycle tension test are plotted in Fig 4 The actual reduction of area values obtained at each temperature are shown along with curves depicting the maxi-mum and minimum limits of ductility

As can be noted, neither steel exhibited any appreciable ductility at 2550 F This was as expected because this tem-perature is only slightly lower than the solidus temperature As the test tem-perature was lowered, the ductility in-creased rapidly to quite high values The ductility was highest over the range from approximately 2300 to 2400 F Below 2300 F, the ductility decreased, which is characteristic of steels tested

in this temperature range From these data, it appeared that neither steel was susceptible to weld metal hot crack-ing, which occurs in some materials be-cause of low ductility at temperatures just below the nominal solidus

Discussion:

Direct-arc air melted AISI 316 has been compared with consumable-elec-trode vacuum-arc remelted AISI 316 in

a variety of respects The properties involved in the comparison have been composition, workability, tensile prop-erties, impact values, creep, stress-rup-ture, inclusion content, behavior in the Huey corrosion test, and weldability

In almost all respects, both materials behaved the same and were of equal quality The principal difference in behavior was the performance of sensi-tized specimens in the Huey boiling 65 per cent nitric acid corrosion test And this difference can be ascribed entirely

to the difference in carbon content tween the two heats of steel

be-From the results obtained, the only clearly discernible difference between the two heats of stainless steel was in composition The vacuum-melted steel was lower in carbon, oxygen, and hy-drogen Here, it can be said that the low oxygen and low hydrogen were a particular consequence of the melting method However, no performance or quality effect, as measured in terms of the parameters used in this investiga-tion, could be attributed to low oxygen

or low hydrogen or to any other factor specific to the method of melting In

ADVANCES IX THE TECHNOLOGY OF STAINLESS STEELS

sum, since only minor differences were

found between the air-melted and the

consumable-electrode vacuum-arc

re-melted material, and these differences

were not found to influence quality and

performance, no technical basis has

composition of the steels is given in Table 10; the first two were reported as having low tensile ductility, while the third was said to have normal ductility

From Material A, 12 specimens, 3

by 2^ in., were sheared for Olsen cup

emerged for melting AISI 316 stainless

steel by practices other than standard

air-melting procedures

VACUUM ANNEALING

Materials and Procedures:

The experimental materials comprised

cold rolled and annealed 0.020-in.-thick

by 3-in.-wide stock from three different

heats of AISI 301 stainless steel The

testing Twelve longitudinal strip sion specimens with reduced-section dimensions of 5 by 2 | in were machined from Material B From Material C, 12 cup and 12 tension specimens were prepared

ten-Four cup specimens from Materials

A and C and four tension specimens from Materials B and C were annealed under each of the following conditions:

1 in air at 1900 F for 30 min

2 in vacuum at 1900 F for 30 min

3 in hydrogen at 1900 F for 30 min The specimens annealed in air were cooled in air For annealing in hydrogen,

a horizontal furnace with an Inconel muffle and a hydrogen flow rate of 8 to

10 cu ft per hr was employed This muffle was so constructed that one end

of it extended outside the furnace and was water cooled Annealing was ac-complished by placing the specimens in the cold zone of the muffle, purging the muffle with hydrogen, and then moving the specimens into the hot zone of the muffle After annealing for 30 min, the specimens were moved to the cold zone where they cooled in hydrogen

For vacuum annealing, a vertical furnace was used This furnace was con-structed with a "bell" section at the top which operated in the vacuum, yet was maintained at about room tempera-ture Annealing was accomplished by inserting the specimens, held in a specially constructed rack, in the bell section A vacuum was then drawn and the specimens were lowered into the hot zone Upon annealing for 30 min, the specimens were returned to the bell section where they cooled in the vacuum

The equipment used developed a vacuum

of 5 X 10~* mm of mercury; however, upon placing the specimens in the hot zone, the vacuum dropped to 0.25 mm

About 15 min was required to regain the high vacuum of 5 X 10~* mm It was

noted that, during vacuum annealing,

a white powdery deposit had formed

on the inside of the furnace This posit may possibly have been chromium which had vaporized from the specimens and condensed on the furnace parts during annealing

de-Strip specimens for hydrogen nations were annealed with the cup and tension specimens under the three condi-tions

determi-Results:

The results of hydrogen tions on material annealed in the three different environments are given in Table 11 After each annealing treat-ment, all three materials had very similar hydrogen contents Those annealed in air had about 1.2 ppm, those annealed

determina-in vacuum had about 0.5 ppm, and those annealed in hydrogen contained about

8 ppm hydrogen

The results of the tension tests are given in Table 12 For Material C of normal ductility, annealing in air and in vacuum resulted in average tensile elongations of 85 and 81 per cent, respectively, while annealing in hydrogen produced a ductility of 36 per cent For Material B with low ductility, the cor-responding elongation values were 47.5 42.5, and 24 per cent Annealing in air

in hydrogen, or in vacuum had little effect on the yield strength of either ma-terial However, annealing in hydrogen resulted in a noticeable decrease in the ultimate tensile strength of both The results of the Olsen cup tests are given in Table 13 For Material C

of normal ductihty, depths of tion at failure of 0.493, 0.526, and 0.384

penetra-in after anneahng in air, in vacuum, and

in hydrogen, respectively, were obtained For the low-ductility stock Material A the corresponding values were 0.478, 0.470, and 0.397 in For Material C, loads of 4600 and 5000 lb were required

to produce failure of the material ai: annealed and vacuum annealed, while a load of only 3100 lb was observed for the specimens annealed in hydrogen Foi Material A, no significant difference in the load at failure (4800 lb) was noted for specimens vacuum annealed or aii annealed For material hydrogen an-nealed, the load at failure was only

3400 lb

Ends of the cup and tension specimens which had been annealed in hydrogen and in vacuum were examined metal-lographically A slight amount of grain-boundary carbide precipitation was

ROACH ET AL ON VACUUM MELTING AND ANNEALING

observed by use of the electrolytic oxalic

acid etch The amount of carbide

pre-cipitation was of the same order for

specimens annealed in hydrogen as it

was for those annealed in vacuum

Ma-terial C appeared to show somewhat

more carbide precipitation than the

others, possibly because of its somewhat

higher carbon content The specimens

annealed in hydrogen had a somewhat

larger grain size than those annealed

in vacuum

Discussion of Results:

The results of this study indicate that

vacuum annealing at 1900 F for 30

min can result in a low hydrogen

con-tent (0.5 ppm) in Type 301 stainless

steel strip However, annealing in air

also can result in a low hydrogen content

(1.2 ppm) Annealing in hydrogen, on

the other hand, can produce a

considera-ble increase in the hydrogen content of

the steel

No significant differences in yield

strength, tensile strength, or cup or tensile ductility were found in specimens annealed in vacuum and in air Annealing

in hydrogen, however, resulted in (1)

a drop in tensile ductility of about 50 per cent, (2) a reduction in cup ductility

of about 20 per cent, (3) a reduction in tensile strength of about 25 per cent, but no significant change in yield strength It is inferred that the increase

in hydrogen content which occurred on annealing in hydrogen was responsible for these changes Similar effects have been observed in the commercial bright annealing of AISI Type 301 strip in hydrogen-containing atmospheres How-ever, the magnitude of such effects has been much smaller because the time

at temperature in commercial bright annealing is much shorter

A cknowledgment:

The authors wish to express their appreciation to the Sharon Steel Corp

for sponsoring the programs upon which

this paper is based At the same time, the authors wish to thank the company for granting permission to publish the results reported here

REFERENCES (1) J H Moore, "Promises and Problems Posed

by Vacuum Melting," Journal of Metals,

Vol 6, 1954, pp 1368-1369

(2) J H Moore, "Vacuum-Induction Melting,"

National Symposium on Vacuum nology Transactions, Edited by E A Perry

Tech-and J H Durant, Pergamon Press, New York, N Y., 1956, pp 202-208.,

(3) K Bungardt and H Sychrovsky, erties of Steels Melted Under Reduced

"Prop-Pressure," Stahl und Eisen, Vol 76, No 16,

1956, pp 1040-1049

(4) W W Dyrkacz, "Quality Improvements

in Stainless Steels and Superalloys by

Vac-uum Melting," Metal Progress, Vol 75, 1959,

pp 138, 140-142, 144

(5) R P Sopber, A J Jacobs, and P J Rieppel,

"Investigation of Weld-Metal Cracking in

High-Strength Steel," The Welding Journal,

Vol 34, No 11, 1955, Research ment, pp.544S-552S

AustroUing Crucible 422 stainless steel increased yield strength by as much

as 45 per cent and tensile strength by 30 per cent The fracture toughness

characteristics, as indicated by net fracture strength and K^ values, were

significantly superior after austroJling The net fracture strength of

conven-tionally heat-treated steel was 122,000 psi and of austrolled steel was as high

as 212,000 psi

The increased resistance to cracking under stress in a salt solution and in a

cathodically hydrogen-charged environment (hydrogen cracking) was

sig-nificantly improved after austrolling

A relationship was shown between mechanical fatigue notch crack

propaga-tion and stress corrosion and hydrogen cracking The microstructural changes

imparted by austrolling influenced fracture characteristics that were reflected

in mechanical and electrochemical properties of the steel

Metastable austenite deformation is a

strengthening technique for hardenable

steel that involves plastic deformation

below the recrystallization temperature

as well as in a temperature range where

no decomposition of austenite occurs

On subsequent cooling, the deformed

austenite transforms to martensite or

bainite or both Metastable austenite

deformation has also'been described as

ausworking, ausforming, and austrolling

The early reports ,(l ,2)- on this process

prompted several investigations (3,4,5)

Although some research has been

con-ducted on the effect of metastable

austenite deformation on the properties

of 12 per cent chromium stainless steels

(6-10), only limited data are available

on the high-strength modified 12 per

cent chromium steels, that is, those with

secondary hardening characteristics The

present work was undertaken to provide

a better understanding of the mechanism

of austrolling and to evaluate the effect

of this process on directionality of tensile

properties, notched tensile properties,

tempering characteristics, and

stress-corrosion and hydrogen-cracking

charac-teristics of such a steel

MATERIALS

Crucible 422—a 12 per cent chromium,

hardenable stainless steel—was ideally

suited for austrolling since an extremely

^ Former staff metallurgist and director of

product research and development,

respec-tively, Crucible Steel Company of America,

Pittsburgh, Pa

^ The boldface numbers in parentheses refer

to the list of references appended to this paper

long incubation period exists in the bay area of the time-temperature-transforma- tion curve, and no bainite reaction takes place (Fig 1) Air-melted material from

a commercial heat (30-ton) and induction-melted material from a labora- tory heat (50-lb) were used in this investigation The air-melted steel was

vacuum-in the form of 0.090-vacuum-in.-thick sheet, and

Ahe vacuupi-melted steel in the form of

0.75-in.-thick.plate The composition of the air-melted heat conformed with the nominal composition of Type 422; how- ever, the carbon and nitrogen contents

of the vacuum-melted heat were lower than those normally found (Table 1)

PROCEDURE

Austrolling:

To secure adequate reductions by rolHng as well as to minimize heat losses, sandwich composites (4 by 6 in.) of three or four of the 0.090-in.-thick sheets

of the air-melted steel were prepared by welding all four edges To achieve higher reductions in thickness than were possible with sandwich composites, 2-in.-long, 3-in.-wide samples of the vacuum-melted 0.75-in.-thick plate were used

Rolling was done on a two-high tory mill The sandwich composites and the plate samples w^ere austenitized at

labora-1900 F for 30 min and 1 hr, respectively, air-cooled to approximately 100 F above the desired rolhng temperature, charged into a furnace held at the rolling tem- perature, soaked for 10 to 15 min, and

rolled Two temperatures, 1000 and

1150 F, were chosen for deformation temperatures, and reductions in thick- ness up to 90 per cent were achieved The sandwich composites were reheated be- tween every two passes, and the plates were reheated after each pass Through the use of Tempilsticks, the estimated lowest temperatures encountered during rolling at 1000 and 1150 F were ap- proximately 900 and 1000 F, respec- tively After being rolled to the desired thickness reductions, the sandwich com- posites were air-cooled, and the plate samples were oil-quenched to room tem- perature

For comparison purposes, thick samples of the air-melted steel and 0.080-in.-thick samples (hot-rolled from the 0.75-in.-thick plate) of the vacuum-melted steel were convention- ally hardened by austenitizing a t 1900 F for ^ and 1 hr, respectively, and oil- quenched

0.090-in.-With the exception of the material used in tempering studies, all samples were tempered at 980 F for 1 hr Only the center one or two sheets were taken from the sandwich composites for testing purposes To eliminate the effects of oxidation and decarburization, all sam- ples were surface ground

Tension Tests:

Unnotched tension-test specimens in both the longitudinal and transverse directions were prepared from the con- ventionally heat-treated and the aus- trolled samples Two types of longitudi- nal notched test specimens, edge-notched and center-notched (Fig 2), were also prepared For the edge-notched speci- mens, the root radii were ground so that the theoretical stress concentration

factor (Kt) was 6.3 The center-notched

specimens were prepared in accordance with ASTM recommendations (11); the notches were made by electrical dis- charge machining followed by cracking

in a tension-tension fatigue machine Both types of notched specimens were tested at a crosshead travel speed of 0.05 in./min

Copyright® 1965 by ASTM Intemational

Except for the testing on unnotched

longitudinal specimens of the

vacuum-melted steel at 1000 F, all tension tests

were conducted a t room temperature

Tempering Studies:

To study the effect of austroUing on

the tempering characteristics of Type

422, conventionally hardened and

aus-trolled samples of the vacuum-melted

5.00 in long) were prepared from both conventionally heat-treated and aus- trolled samples of the vacuum-melted steel A\l specimens were stressed to 150,000 psi by bending (2-point load- ing) in test fixtures and then finished with No 360 grit paper The specimens were coated with acrylic resin—except for the area of maximum stress—to eliminate any secondary galvanic effects

5 10 20 40

Minutet Time

2 S 10 20

>teurs

F I G 1—Time-Temperature-Transformation Diagram for Crucible 422 Stainless Steel

TABLE 1—CHEMICAL COMPOSITIONS OF CRUCIBLE 422

STAINLESS STEEL, P E l i CENT

Mo

0 9 4 1.13

steel were tempered for 1 hr at selected

temperatures in the range of 800 to

1200 F Hardnesses (DPH) were

de-termined by taking an average of seven

readings on transverse sections of

sam-ples with a Tukon testing machine and

subsequently converting D P H numbers

to those on the Rockwell C scale

Stress-Corrosion and Hydrogen-Cracking

Tests:

Test specimens (0.50 in wide and

One series of tests was performed in

of cracks when the surfaces were viewed

solu-R E S U L T S AND DISCUSSION

Tension Tests:

The results of the room-temperature tension tests on the air-melted steel (Table 2 and Fig 3) showed t h a t the difference in austrolling temperatures (1150 and 1000 F) had little effect on tensile properties

The longitudinal tensile strength creased continually with increasing deformation up to a thickness reduction

in-of 75 per cent; higher reductions could not be achieved because of limitations posed by sample size and rolling equip- ment However, the yield strength was not significantly increased until thick- ness reductions in excess of 60 per cent had been obtained Comparison of edge- notched and unnotched tensile strength values showed that austroUing decreased the notch sensitivity of Type 422 I n general, the elongation tended to de- crease with increasing deformation The transverse yield and tensile strengths • of the air-melted steel in- creased more rapidly than the longitudi- nal strengths with increasing deforma- tion, and, at 75 per cent reduction in thickness, the directionality of strength properties was marked Although the transverse elongation values decreased with increasing deformation, the de- crease was not as great as for the longi- tudinal elongation

The results of room-temperature sion tests on the vacuum-melted steel (Table 3) showed that sizable increases

ten-in both the longituciten-inal and transverse yield and tensile strengths were effected

by austrolling to thickness reductions of

85 to 90 per cent Although the tion decreased as a result of deformation

elonga-in the metastable austenite temperature range, the resultant ductility was still quite reasonable—transverse and longi- tudinal elongations were decreased the same amount There was no indication

of notch sensitivity at a Kt of 6.3 The

magnitude of the change in, the tudinal room-temperature strength prop- erties imparted by austrolling may be summarized as follows:

longi-10 ADVANCES IN THE TECHNOLOGY OF STAINLESS STEELS

o r s o D M Hot*

« 002 R M R I T Y P )

Change Due to Austrolling, per cent

85 per cent 90 per cent

at 1000 F at 1150 F

FIG 2—Notched Tension-Test Specimens

TABLE 2—ROOM-TEMPERATURE TENSILE PROPERTIES OF T H E

t r a n s

t r a n s long

long

t r a n s

t r a n s

0.2 Per Cent Offset Yield Strength, 1000 psi

These results indicate that, for a given deformation, higher strength is obtained

at lower austrolling temperatures

The results of longitudinal tension tests conducted at 1000 F (Table 3) showed that austrolling also improved the elevated-temperature strength of the vacuum-melted steel This improvement was most pronounced for the material austrolled 85 per cent at 1000 F for which increases of 20 and 14 per cent in yield and tensile strengths, respectively, were obtained

It should be noted that the hardness and hence the tensile strength of the conventionally hardened and tempered vacuum-melted steel were considerably lower than the hardness and tensile strength of the conventionally hardened and tempered air-melted steel This discrepancy is probably due to the difference in the carbon and nitrogen contents of the two heats (Table 1) The as-quenched and quenched and tempered hardness values for the two heats differed by as much as 4 points Rockwell C

The results of center-notched tension tests (Table 4) showed that the fracture toughness of the vacuum-melted steel was considerably improved by austrol-ling On the basis of net fracture strength,

a greater improvement in fracture ness was effected by austrolling at 1000 F than at 1150 F; the average net fracture strength values for the conventionally hardened and the 1000 F austrolled materials were 122,000 and 212,000 psi, respectively Although the average ratios

tough-of net fracture to yield strength tough-of the conventionally hardened and 1150 F austrolled materials were similar (0.72 and 0.68), this ratio was considerably higher (0.86) for the 1000 F austrolled material When the three materials are

compared on the basis of K^, another

fracture toughness parameter, no ences exist between the materials aus-trolled at 1000 and 1150 F; however, the austrolled materials were superior to the conventionally heat-treated ma-

differ-terial The agreement between the Kd

values (based on critical crack length measured by 24-frames/sec moving pic-

tures) and the Kcs values (based on

WESTGREN AND DULIS ON THE EFFECTS OF AUSTROLLING 11

c _o

0 20 40 60 80

Reduction in Thickness by Austrolling (%)

F I G 3—Effect of Austrolling a t 1000 a n d 1150 F on t h e R o o m - T e m p e r a t u r e Tensile P r o p e r t i e s

Tem-Direction of Test

0.2 Per Cent

Ofiset Yield

Strength, 1000 psi

Tensile Strength,

1000 psi Unnotched Notched''

long, long

long, long

long, long

10.0 7.5

6.0 5.5

" 411 samples were a u s t e n i t i z e d a t 1900 F for 1 hr, h a r d e n e d a s i n d i c a t e d , a n d t h e n t e m p e r e d

a t 980 F for 1 h r

''K, = 6.3

measurements of percentage shear) was

very good; unfortunately, K^ values

could not be determined for the 1000 F austrolled material because the scope

of the available plot would not permit the graphical solution of the essential

q^ values

The difference in fracture appearance

of the austrolled and conventionally hardened center-notch test specimens was quite striking (Fig 4) The average percentage shear values (measured at lOX) for 1150 F austrolled, 1000 F austrolled, and conventionally hardened test specimens were 77, 75, and 23 per cent, respectively

Metallographic examination of the fracture regions of the specimens showed

a difference in fracture mode The ventionally heat-treated specimen pri-marily exhibited both the large angular contour and the small grain deformation typical of brittle-type fracture, whereas the austrolled specimens exhibited both

con-a smooth contour con-and grcon-ain elongcon-ation typical of ductile-type fracture

Electron microfractographs' (Fig 5)

of the center portions of the fractured center-notched specimens (Fig 4) clearly distinguish between the modes of frac-ture The conventionally heat-treated

steel fractures (Fig 5 (top)) show the

characteristics of brittle fracture such

as large cleavage facets containing

"river" marks {A) and cleavage

"tongues" (B) The austrolled steel fracture (Fig 5 {bottom)) shows the

randomly oriented "dimpled" structure characteristic of flat transverse portions

of ductile fracture

The increased fracture ductihty of the austrolled steel is believed to be attributable to the decrease in size and more uniform distribution of both pre-cipitate and dislocation substructure in the austrolled steel; these are factors that would reduce dislocation pileups and, hence, would reduce the probability

of fracture Distortion of the martensite during the austenite-to-martensite trans-formation occurs due to the martensitic growth process in the deformed (aus-trolled) austenite, wherein barriers of deformation bands in the austenite have

to be overcome (13) Upon tempering at

980 F for 1 hr, the following processes take place in an interrelated manner:

' A n electron microfractograph t e c h n i q u e involving a s t r i p p e d cellulose a c e t a t e replica was used T h e replicas s t r i p p e d from t h e frac-

t u r e surface were m e t a l s h a d o w e d b y a r o t a r y

t e c h n i q u e a n d c a r b o n coated T h e c a r b o n replica was s e p a r a t e d from t h e p r i m a r y a c e t a t e replica b y dissolving t h e a c e t a t e film in w a r m

12 ADVANCES IN THE TECHNOLOGY OF STAINLESS STEELS

(1) carbide precipitation, (2)

polygoniza-tion in the martensite, and (3) partial

relief of residual stresses In contrast,

conventionally heat-treated steel

con-tains well-defined undistorted martensite

needles developed during the

austenite-to-martensite transformation Upon

tem-pering, carbide precipitation and some

stress relief occur Thus, the increase in

fracture toughness of the austrolled

imparted by austroUing is directly associated with higher strengths at elevated temperatures, and this behavior

is shown by the increase in yield and tensile strength of the austrolled steel

at 1000 F

Representative quenched and ed-and-tempered microstructures of the conventionally hardened and the aus-trolled materials (Fig 7) show that a TABLE 4—ROOM-TEMPERATURE FRACTURE TOI'GHXESS CHARACTERISTICS

quench-OF THE VACUUM-MELTED STEEL."

" 0.045-in.-thick longitudinal specimens

''All samples were austenitized at 1900 F for 1 hr, hardened as indicated, and then tempered at

980 F for 1 hr

(top) Conventionally hardened (bottom) Austrolled at 1150 F

FIG 4—Fracture appearance of Center-Notched Tension-Test Specimens

steel may be associated with the

signifi-cantly smaller fracture facets in steel

given this prior treatment as compared

with conventionally heat-treated steel

Tempering Studies:

The results of the tempering tests on

the vacuum-melted steel showed that,

compared to conventionally heat-treated

material, the hardness increase resulting

from austroUing was not only maintained

but increased after tempering (Table 5

and Fig 6) The retarded tempering

highly distorted martensitic ture was inherited from the heavily deformed austenite A detailed explana-tion of the mechanisms involved in the tempering of this steel after austroUing has been included in another publication

microstruc-(13)

Stress-Corrosion Tests:

and Hydrogen-Cracking

For martensitic 12 per cent chromium steels (including the Type 422 of the present investigation), the usual form of

cracking is hydrogen cracking The more common form of stress-corrosion crack-ing associates (1) crack formation at anodic reaction sites and (2) crack propagation by either a continuous electrochemical mechanism or a periodic electrochemical-mechanical mechanism (14) Hydrogen cracking is not initiated

at anodic reaction sites

The results of the stress-corrosion tests performed in the 5 per cent NaCl solution under a stress of 150,000 psi (Table 6) showed that austroUing signifi-cantly increased time to failure Also,

in the cathodic hydrogen cracking tests, similar results were found As shown in Figs 8 and 9 for the hydrogen crack test specimens, the cracks that formed

in the austrolled materials were short and discontinuous, whereas those that formed in the conventionally hardened material were longer with more con-tinuous paths The explanation previ-ously given in relation to fracture tough-ness in tensile stressing also applied to fracture mode in stress corrosion and hydrogen cracking The shorter fracture paths of austrolled steel are associated with increased resistance to crack prop-agation

The results of these tests show that the significant strength increases im-parted by austroUing do not affect the crack propagation characteristics ad-versely in 12 per cent chromium marten-sitic steels This finding is in agreement with the results of the center-notched tensile fracture toughness properties previously discussed

CONCLUSIONS

1 The tensile strength of Type 422 stainless steel increased continuaUy with increasing deformation by austroUing; significant increases in yield strength were not obtained until reductions in thickness in excess of 60 per cent were obtained

2 AustroUing resulted in ality of strength properties; the trans-verse yield and tensile strengths were higher than the longitudinal yield and tensile strengths

direction-3 Tensile ductiUty (elongation) creased with increasing deformation

de-WESTGREN AND DULIS ON THE ErrECTS or AUSTROLLING 13

F I G

5-((op) Conventionall\- hardened (bottom) Austrolled at 1150 F -Electron Microfractographs of Fractured Center-Notched Tension-Test Specimens

TABLE 5—TE.MPERIXG ISTICS OF THE VACUUM-MELTED

29

trolled 90% a t

1

r —

1 A.Q 800 900 1000' 1100

Tempering Tennperoture CF) (Time = I hr)

1200

FIG 6—Variation of Hardness with Tempering Temperature for the Vacuum-Melted Steel

(a) Conventionally hardened; as-quenched (b) Conventionally hardened; tempered at 980 F

(c) AustroUed at 1150 F; as-quenched

(d) AustroUed at 1150 F; tempered at 980 F

FIG 7—Representative Mictostructures of As-Quenched and Quenched-and-Tempered Melted Steel Etched in Picral-HCl (XIOOO)

Vacuum-WESTGREN AND DULIS ON THE EFFECTS OF AUSTROLLING 15 [TABLE 6—RESULTS OF STRESS-COR-

llOSION AND HYDROGEN-CRACKING

TESTS ON THE VACUUM-MELTED

% NaCI Solution, days

mg As per liter, 0.6 amp per sq m., min

" All samples were austenitized at 1900 F

for 1 hr, hardened as indicated, and then

in-800 to 1200 F

8 Austrolling increased the resistance

to cracking under stress in a salt solution and in a cathodically charged environ-ment

REFERENCES

(2)

(top) Conventional!}' hardened

(bottom) Austrolled at UoO F

FIG 8—Typical Appearance of Specimens

That Failed in the Hydrogen-Cracking Test

(1) R F Harvey, "Step Quenching, Hot

Peening, Improve Lean Alloys," Iron Age,

Dec 27, 1951, Vol 168, pp 70-71

E M H Lips and H van Zuilen,

"Im-proved Hardening Technique," Metal Progress, August, 1954, Vol 66, pp 103-

104

(3) C W Marschall, "Hot-Cold Working of

Steel to Improve Strength," Report No 192,

Defense Metals Information Center, Oct

11, 1963

(4) V F Zackay, M W Justusson, and D J

Schmatz, "Strengthening by Martensitic

Transformations," Strengthening

Mecha-nism in Solids, Am Soc Metals, 1962, pp

(6) F A Malagari, "Discussion to Paper by

J C Shyne, V F Zackay, and D J

Schmatz," Transactions, Am Soc Metals,

Gar-Journal Iron and Steel Inst., Vol 196,

September, 1960, pp 66-81

(9) Y Hosoi and K Pinnow, "The Tensile Properties of Type 410 Stainless Deformed Before and After Martensite Transforma-

tion," Transactions, Am Soc Metals, Vol

TjT^e 422 Stainless Steel," Transactions,

Am Soc Metals, Vol 56, 1963, pp

629-642

(14) S Barnartt, "General Concepts of

Stress-Corrosion Cracking," Stress-Corrosion, Vol 18,

September, 1962, pp 322t-331t

DISCUSSION

J J HEGER'—The austrolled steels increase the fatigue and corrosion fatigue martensitic steels Unfortunately, oui and conventionally heat-treated steels strength to the same proportion that it tests were limited to those reported ir were tested in stress corrosion at the increases the tensile and yield strength? the paper, so we are unable to shec same stress level What would be the R C. WESTGREN A.ND E J. DULIS further light on the questions raised

results if the stress were a given fraction (authors)—The points raised by Mr Additional work will be required to morf

of the yield strength? Does austroUing Heger are pertinent for a fuller under- fully resolve such aspects as fatigue

> Chief research engineer, stainless steel, standing of the eflfects of austrolling on corrosion fatigue, and fracture tim<

U S Steel Corp., Pittsburgh, Pa the properties of 12 per cent chromium under different stresses

16

A series of nickel-chromium stainless alloys made by air and vacuum

melt-ing were rolled 20 or 40 per cent at temperatures rangmelt-ing from —320 to 400 F

Smooth and sharp-notched sheet-tensile properties were measured after heat

treating the cold-rolled pieces for 24 hr at 800 F Yield strengths up to 260 ksi

were obtained Hardening resulted primarily from martensite formation, and

also from work hardening of the austenite prior to transformation The yield

strengths were correlated in terms of composition

Lowering the silicon content to 0.1 per cent significantly raised the

tough-ness Over the ranges studied none of the remaining elements had effects

com-parable to silicon The toughness at a given strength level also depended upon

the rolling conditions, but not upon the per cent martensite

Tests at -320 F with a low-silicon alloy showed a notch-tensile ratio of 1.0

at 297 ksi ultimate tensile strength The results suggest that a low-silicon

301-type composition would have good cryogenic properties

Several investigations have shown

t h a t very high strengths can be obtained

by cold or sub-zero roUing Type 18-8

stainless steels (l-ll).' To date, however,

there has been little systematic study of

the effects of composition on the

mechan-ical properties of stainless steels treated

this way I t is recognized that lower

alloy content grades are more easily

hardened by cold rolling or sub-zero

rolling, but information on the influence

of roUing conditions and composition on

strength properties is lacking One of the

objects of the present program was to

survey the influence of composition and

rolling conditions on strength

A more serious problem is t h a t of

toughness The general results have

indicated that when the yield strengths

of cold-rolled stainless steels are on the

order of 25PJPQj3si, toughness is poor

There is some evidence that lower carbon

and nitrogen contents improve

tough-ness (4, 5, 10) There apparently has

been no consistent study, however, of

the influence of alloying elements on

toughness A study of toughness versus

composition was the second object of this

' T h e boldface n u m b e r s in parentheses refer

t o t h e list of references a p p e n d e d t o t h i s paper

EXPERIMENTAL PROCEDURE Only variations in the principal alloying elements, nickel, chromium, manganese, sihcon, carbon and nitrogen, common to stainless steels, were made

in this study No attempt was made to evaluate the effects of subversive ele- ments, or other alloying elements such

as titanium or columbium which are used

in some grades of stainless steel The carbon contents in this study were in the 0.04 to 0.08 weight per cent range, and the nitrogen contents varied from 0.01

to 0.05 per cent The remaining ments varied approximately as follows:

ele-5 to 10 per cent nickel, 17 to 20 per cent chromium, 0.01 to 1.0 per cent manganese, and 0.01 to 0.5 per cent sihcon

Thirty-pound heats were prepared using electrolytic metal charges Heats were made by either air induction melt- ing or vacuum induction melting The air-melted heats containing less than 0.14 per cent silicon were deoxidized by adding 0.1 per cent aluminum The final aluminum contents in these heats were on the order of 0.05 per cent The compositions of all the heats are given

in Table 1

The 4 by 4 in ingots from each heat were soaked one hour a t 2100 F, rough forged, cropped, and then forged to 1-in thick plate The plates were soaked

1 hr at 1900 F and hot rolled to | - i n thicknesses They were then cold rolled,

17

www.astm.org

annealed for 1 hr a t 1950 F, and air cooled The sheets were then given a final light machining to remove scale and to obtain the desired thickness prior to the final rolling

Sub-zero rolling was done primarily

a t —106 F to a total reduction of ness of 40 per cent A number of addi- tional specimens were also rolled 20 per cent at —106 F, and 20 or 40 per cent a t

thick-— 320 F Also some specimens were rolled

20 or 40 per cent at -+-70 F, + 2 0 0 F, and + 4 0 0 F These latter tests were made primarily to evaluate some of the effects of rolling temperature on the strength

Rolling was done using a two-high Stanat mill with 8-in diameter rolls The reductions per pass were on the order

of 0.005 to 0.008 in except for the final pass, which was generally lighter All of the sheets were rolled to a final thickness

of 0.060 in

Baths of solid CO2 plus isopentane ( - 1 0 6 F ) , and hquid nitrogen ( - 3 2 0 F) were used for the sub-zero rolling Strips were cooled in the bath, removed, and given one pass through the rolling mill, and then returned to the bath and allowed to come to the bath temperature again before making another pass

A water b a t h was used for the + 7 0 F rolling, and a small electric oven for rolling a t + 2 0 0 F and + 4 0 0 F The general procedure was the same as t h a t used for sub-zero rolling

Although the strips before rolling were initially a t some known temperature, the actual temperature during deformation may have been considerably different from this initial value Brown and co- workers (9) have attempted to estimate the sheet temperature during rolling by measuring the temperature immediately before and immediately after rolling In the present work, the initial temperature prior to roUing is designated as the roll- ing temperature, without attempting to estimate the exact temperature during rolling This simplification is reasonable for the present comparative type of tests

18 ADVANCES IN THE TECHNOLOGY OF STAINLESS STEELS

After rolling, the specimens were

heat treated for 24 hr at 800 F Heat

treatments of this type raise the strength

of sub-zero rolled 18-8 alloys (1-3, 6)

The reasons for the resultant benefits

are not certain This particular heat

treatment is not necessarily the optimum one, but it was beyond the scope of this work to study this variable in detail A few specimens, however, were tested in the as-rolled condition without any heat treatment

Alloy Yield Strength, Tensile Strength in 2 in., Strength (NTS),

ksi (UTS),ksi % ksi

Net Fracture Stress, ksi NTS'UTS

0 9 4

0 9 2 1.00 1.02 1.09 1.05

0 9 7 1.01

0 8 8

0 8 9

1.08 1.18 1.10 1.00 1.05

The strips were then machined to NASA-type smooth and edge-notched tension specimens (4) The notched tensile specimens had a stress concentra-tion factor, X j , of 18 or larger

The majority of the tension tests were conducted at room temperature Ink staining was used to estimate the net fracture stress (12) Several specimens were tested at —320 F using a liquid nitrogen bath around the specimens Only longitudinal specimens were tested, because the experimental rolling mill did not have sufficient capacity to roll transverse specimens For reductions

up to 40 per cent, however, the cal anisotropy of cold-rolled stainless steel is not severe (3,8,9), and thus longitudinal specimens were considered

mechani-to be satisfacmechani-tory for screening tests

A number of specimens were examined

by X-ray diffraction to determine the phases present and the amounts of martensite formed Many specimens were also examined by optical and elec-tron microscopy The resistance of two sub-zero rolled alloys to atmospheric stress corrosion was also evaluated The results of these latter studies will also be described

RESULTS AND DISCUSSION

Ejects of Rolling Conditions on Strength:

Yield-Table 2 summarizes the perature tensile properties of the speci-mens heat treated 24 hr at 800 F after rolling The results show that yield strengths between 68,000 and 275,000 psi were obtained with changes in rolling temperature, reduction, and composition Figure 1 shows the yield strength of three of the alloys versus rolling tempera-ture after 20 per cent reduction The least stable alloy, containing 5.3 per cent nickel, had an essentially constant yield strength of 195,000 psi over the entire range of rolling temperatures The most stable alloy, containing 10.1 per cent nickel, had relatively low strengths over this range The intermediate alloy, containing 7.6 per cent nickel, had low yield strengths after warm rolling, but showed a marked increase in strength with sub-zero rolling

room-tem-Figure 2 compares the effects of 20 and

40 per cent reduction on the yield strengths of two of these alloys The

5.3 per cent nickel alloy showed an

in-crease in yield strength of only about 10,000 psi with 40 per cent reduction compared to 20 per cent reduction

FLOREEN AND MIHALISIN OX LOW-TEMPERATURE ROLLING 19 Conversely, much larger gains in yield

strength were obtained with the 10.1

per cent nickel alloy using 40 per cent

reduction

The strengthening of 18-8 stainless steel by cold work results primarily from the transformation of metastable austen-ite to martensite during deformation

cold-100 per cent martensite Figure 3 lustrates the relation between yield strength and martensite content as measured by X-ray diffraction for the 10.1 per cent nickel alloy Figure 4 is an optical microscopic study of the micro-structure of this 10.1 per cent nickel alloy after the two reductions (20 and

il-40 per cent) at various temperatures There is a marked difference between the structures obtained upon rolling at sub-zero and elevated temperatures The structures obtained upon rolling at elevated temperatures appear to bear only slip traces from deformation whereas the structures formed at sub-zero temperatures are darkly etching and reveal orientation relationships between the martensite and parent austenite lattice

No microstructural difference could be observed here to account for the different strength levels at a given martensite content as shown in Fig 3 For instance, rolling 20 per cent at —320 F results in the same amount of martensite forma-tion as rolling 40 per cent at —106 F Any difference in microstructure is not obvious

Electron micrographs of these same structures delineate the martensite with more clarity This is shown in Fig 5, micrographs of shadowed negative rep-licas after rolling 20 and 40 per cent at

- 1 0 6 F and - 3 2 0 F The structures are similar to those shown by Reed (13) It appears that the martensite needles are parallel to (111) austenite planes Again the effect of degree of rolling on strength as shown in Fig 3 is not obvious from these higher magnifica-tion studies

The epsilon (hexagonal close-packed) transition phase observed by many workers in studies of metastable stainless steels was observed in this work only under special conditions The (10.1) epsilon peak was clearly observed in X-ray diffraction patterns of alloys which transformed partially to mar-tensite on cooling to room temperature

No epsilon phase was observed after cold rolling and heat treating This result is in accord with previous work in which the epsilon phase disappears with large amounts of deformation or heat treatments on the order of 800 F (13-

16)

The main reason for achieving higher

20 ADVANCES IN THE TECHNOLOGY OF STAINLESS STEELS

FIG 1—Yield Strength Versus Rolling Temperature for Three Alloys All Specimens Rolled 20

Per Cent and Heat Treated at 800 F for 24 hr After Rolling

• 2 0 % REDUCTION -300 -200 -100 0 100 200

ROLLING TEMPERATURE ("F)

300 400

FIG 2—Yield Strength Versus Rolling Temperature for Two Alloys Specimens Rolled 20 and 40

Per Cent and Heat Treated at 800 F for 24 hr After Rolling

FIG 3—Yield Strength Versus Per Cent Martensite for Alloy No 2 (10% Xi) Cold Rolled 20 and

40 Per Cent at Various Temperatures Heat Treated at 800 F for 24 hr After Rolling

Strengths exemplified by Figs 1 and 2 is

the formation of martensite by rolHng

With lower rolling temperatures and

larger reductions increasing amounts of

martensite are formed, and the strength

increases accordingly The data for the

7.6 per cent nickel alloy in Fig 1, and the 10.1 per cent nickel alloy in Fig 2 suggest that most of the strength change

is produced within a relatively small change in rolling conditions

There appear to be other factors

in-fluencing strength in addition to mai tensite formation Figure 3, for example shows that for a given martensit content, the specimens rolled 40 pe cent were significantly stronger tha: those rolled 20 per cent Thus th amount of cold work also is important Presumably this effect is due to wor' hardening of the austenite prior t transformation, or work hardening o the martensite after transformation, o both The results with the 5.3 per cen nickel alloy illustrated in Fig 2 indicat that there is little work hardening c the martensite With this alloy, approxi mately 99 per cent martensite wa formed after only 20 per cent reductioi over the entire range of rolling tempera tures As noted previously, the 40 pe cent reduction increased the strengt] about only 10,000 psi, indicating ven little strengthening from cold rolling o martensite

One must conclude, therefore, tha work hardening of the austenite prior t( transformation produces a significan amount of strengthening which i: incorporated in some unknown fashioi into the final strength of the martensiti (14) This same effect has also beer noted in the related method of strength ening by ausforming (17,18)

Effects of Composition on Yield Strength

Assuming martensite formation to b( mainly responsible for strength, it shoulc

be possible to correlate the effects o:

composition in terms of the Md ture Since the Md temperature is

tempera-proportional to the more easily estimatec

M, temperature (19,20), the correlations

made here will be in terms of M,

The effects of the individual alloying

elements on the Ms temperature have

been estimated by Eichelman and Huf (21) Based on their findings, the relative

stability, S, of an alloy may be defined

as:

.S- = Xi -I- 0.68(Cr) -I- 0.55(Mn)

+ 0.45(Si) -I- 27(C + X) (1)

Higher values of S indicate an alloy that

would be more resistant to tion to martensite

transforma-Plots of j'ield strength versus S for

several different rolling conditions are shown in Figs 6 and 7 The data may be represented with reasonable accuracy by

a smooth curve for each rolling tion Thus, empirically, the correlation

condi-of strength with S appears useful It

should be noted, however, that such a correlation may only be useful within

FLOREEN AND MIHALISIN ON LOW-TEMPERATURE ROLLING 21 :ertain composition Umits In particular,

arger amounts of carbon and nitrogen

nay strengthen the martensite to an

;xtent that overweighs the change in 5

:5)

The curves in Figs 6 and 7 indicate

Lhat for a given roUing condition there is

in optimum value of 5 which will have

the highest strength Similar results have

been shown by Carlsen and Thomas

using a different composition

correla-tion (11) The value of S producing the

highest strength varies somewhat with

the rolling conditions With lighter

reductions and higher temperatures,

this maximum occurs at lower values of

S This would be expected, since more

stable alloys will not transform as

readily to martensite at higher

tempera-tures and lighter reductions

For a given rolling condition, the drop

in strength with increasing S above the

maximum value is due to decreasing

amounts of martensite formation This

conclusion is supported by the X-ray

diffraction results

The drop in strength below the

maxi-mum with decreasing values of S is

puzzling Originally it was believed that

as the austenite became less stable

martensite would form either on cooling

to the rolling temperature, or during the

initial stages of rolling In either case,

there would be little work hardening

of the austenite, and this strengthening

contribution would not be included in

the final strength However, two tests

in which the heat treatment after rolling

was omitted indicate that 5 was a

function of heat treatment also The

results in Figs 6 and 7 are based on

heat treatments of 24 hr at 800 F after

rolling Two of the less stable alloys

were rolled 40 per cent at —106 F and

tested without heat treatment The

results are given in Table 3 It is seen

that when the heat treatment was

omitted the yield strengths were much

higher

This result is surprising, and contrary

to the usual increase in strength

pro-duced by heat treatment Figure 8

exemplifies the more normal gain in

strength with the 800 F heat treatment

for another alloy having a higher

sta-bility These results show a marked

increase in strength with heat treatment

after sub-zero rolling

X-ray diffraction studies of the

less-stable alloys showed amounts of

mar-tensite on the order of 100 per cent after

rolling, whereas the specimens around the

peaks of the yield strength versus 5

curves had martensite contents on the order of 85 to 95 per cent It thus appears that some minimum amount of austenite may be necessary to obtain strengthen-ing by heat treatment after roUing

Bastien and Dedieu (22) have also noted that when a stainless steel was com-

pletely transformed to martensite by cold working there was a loss in strength with subsequent heat treatment Alloys containing some residual austenite, however, showed an increase in strength after heat treating (22)

Examination of specimens before and

(a) Rolled 20% at - 3 2 0 F, 185 ksi yield (d) Rolled 20% at - 1 0 6 F, 101 ksi yield

strength strength

(b) Rolled 40% at - 3 2 0 F, 254 ksi yield (e) Rolled 40% at - 1 0 6 F , 222 ksi yield

strength strength (c) Rolled 20% at 70 F, 90 ksi yield strength (/) Rolled 20% at 200 F, 73 ksi yield

strength

FIG 4—Microstructures of Alloy No 2 (10.1% Ni After Cold Rolling and Heat Treating 24 hr

at 800 F Picric-Hydrochloric Etch (X 250)

22 ADVANCES IN THE TECHNOLOGY OF STAINLESS STEELS

after heat treatment using X-ray

dif-fraction and electron microscopy

re-vealed only a change in etching response;

no changes in microstructure of phase

content could be detected Thus the

influence of heat treatment on structure

has not been defined In view of the

effect of austenite it would seem,

how-ever, that more than stress-relief is

involved One might postulate that

austenite serves as a sink for atoms

diffusing from martensite during heat

treatment

In view of these effects of heat

treat-ment it is clear that the strength versus 5

curves in Figs 6 and 7 are not

funda-mental in nature Such curves might be

of empirical value in allowing the

pre-diction of strength produced by a given

rolling condition and heat treatment

Effects of Composition and Rolling

Con-ditions on Toughness:

In the majority of the tensile results

in Table 2 the net fracture stress

meas-ured by the ink stain technique was well above the yield strength, and accurate

estimates of Gc and Kc were not possible

These results indicate that at yield strengths of 250,000 psi the values of

Gc were in excess of 1000 in.-lb/in.^, and the values of Kc were in excess of

180 ksi Vin^

A much more distinct, and perhaps more meaningful, evaluation of tough-ness can be made in terms of the ratios of notch to tensile strength At lower strengths the specimens almost in-variably had a ratio of 1.0 or higher

With yield strengths above 230,000 psi, however, the distinct composition effect

of silicon becomes evident

Figure 9 summarizes the yield strengths and notch-tensile ratios at two silicon levels of specimens rolled 40 per cent at —106 F The alloys containing 0.44 to 0.57 per cent sihcon had notch-tensile ratios ranging from 0.88 to 0.93, and an average ratio of 0.91 At the same strength level the specimens

containing 0.01 to 0.13 per cent silicor had notch-tensile ratios ranging frorr 0.96 to 1.05, and an average ratio o: 1.01 Thus lowering the sihcon conteni significantly improved toughness

The effect of silicon appeared to b( about the same in both air- and vacuum melted alloys Statistical analysis of the data indicated that none of the remain-ing elements affected toughness to the extent that silicon did The effect ol silicon in the specimens rolled at —320 P

is not certain, because only one of the higher silicon content aUoys had a yield strength above 225 ksi Therefore

it is not possible to make a good parison of the properties at the twc silicon levels

com-Another general result is that, though many of the specimens rolled

al-20 per cent at —3al-20 F had yield strengths comparable to those rolled 40 per cent at

— 106 F, the notch-tensile ratios of the specimens rolled 20 per cent at —320 F were lower Comparing low silicon heats with yield strengths over 230 ksi, the average notch-tensile ratio of the specimens rolled 20 per cent at —320 F was 0.96, while the average ratio for the specimens rolled 40 per cent at —106 F was 1.01 This result is somewhat surprising, because earlier studies of sub-zero rolling indicated that better ductility resulted from hghter reductions

at lower temperatures (2) This earlier conclusion was based on elongation values, however, which do not always correlate with notch-tensile results Thus,

in some cases, better toughness may result from rolling to larger reductions

at higher temperatures

There was no correlation of toughness with per cent martensite In particular, specimens containing 100 per cent martensite had toughness after the 800 F heat treatment comparable to those containing residual austenite This re-sult is shown in Table 3 Thus retained austenite is not a requirement for good toughness in these alloys

The marked influence of heat ment on toughness is also illustrated in Table 3 A detailed study of the effect of heat treatment on toughness has been conducted by Brown and co-workers (3) As already indicated, the reasons for the changes in properties with heat treatment are not clear

treat-—320 F Tension Tests:

As a further evaluation of the effect

of silicon on toughness, specimens from

a high- and a low-silicon heat were

sub-FLOREEN AND MIHALISIN ON LOW-TEMPERATURE ROLLING 23

zero rolled, heat treated, and then tested

at —320 F The results of these tests are

given in Table 4 The marked increase in

toughness of the low-silicon heat is

clearly evident In particular, the notch

ratio of 1.01 for the low-silicon sample

rolled 40 per cent at —106 F indicates

that this alloy may be useful in cryogenic

applications

The reason for the effect of silicon on

toughness is unknown at this time

Metallographic and X-ray diffraction

comparisons of alloys with high and low

silicon content showed no significant

differences in structure No evidence of

the epsilon phase was detected, and thus

the effect of silicon does not seem to be

related to the presence of this phase

during testing

Corrosion Resistance:

No thorough study of the corrosion

resistance of sub-zero rolled stainless

steels has been made From scattered

tests, however, the corrosion resistance

appeared satisfactory when the steels

contained large amounts of martensite

(1,3,23)

A preliminary evaluation of the

re-sistance to atmospheric stress corrosion

of two of the present alloys was made

Small samples of Heats 2 and 5 (10.1

and 5.3 per cent nickel) were cold rolled

40 per cent at —320 F and heat treated

at 800 F Strips 3.0 by 0.062 by 0.1875 in

were ground from the sheets and loaded

in 3-point bending to an outer-fiber stress

of 200,000 psi After over one year

exposure to the Bayonne atmosphere

there was no evidence of cracking or

rusting in these samples

These results again support the general

conclusion that sub-zero rolled stainless

steels have at least satisfactory corrosion

resistance A more comprehensive study

of the corrosion behavior of these

materials would be very useful, however

GENERAL COMMENTS

The general results indicate that

good strength and toughness qualities

can be obtained in sub-zero rolled

stainless steel by use of a suitable

com-position and rolling procedure In itself

this finding is significant, because in the

past the toughness qualities have been

poor at high strength levels

The best strength and toughness

properties were obtained by rolling

low-silicon alloys 40 per cent at —106

F, and heat treating 24 hr at 800 F

The curves in Fig 6 indicate that for

this particular case the optimum value

(c) Rolled 40% at - 3 2 0 F, 254 ksi yield strength

(d) Rolled 40% at - 1 0 6 F, 222 ksi yield strength

FIG 5—Electron Micrographs of Specimens of Alloy No 2 (10.1% Ni) Showing Structures duced by Rolling and Heat Treating 24 hr at 800 F (X8000)

essen-17.5 per cent chromium, the nickel It is worth noting in this regard that content on the basis of Eq (1) should the nominal composition ranges for then be approximately 7.7 per cent The nickel, chromium, and manganese in

24 ADVANCES IN THE TECHNOLOGV OF STAINI.ESS STEELS

TABLE 3—TENSION TEST RESULTS FOR ALLOYS ROLLED 40% AT - 106 F TESTED

WITH AND WITHOUT HEAT TREATMENT OF 24 HR AT 800 F AFTER ROLLING

0.2%

Offset Yield Strength, ksi

Ultimate Tensile Strength (UTS), ksi

tion in

Elonga-2 in., ej

Notch Tensile Strength (NTS J, ksi

Net Fracture Stress, ksi

FIG 8—Yield Strength Versus Rolling Temperature for AUov No 1 (/.e^c Ni) Rolled 20 Per Cent

and Tested As-Rolled and After Heat Treating 24 hr at 800 F

Type 301 stainless steel are broad A

composition on the low side in these

elements would have a yield strength of

about only 200,000 psi, instead of about

250,000 psi Thus the total composition

should be controlled to somewhat

nar-rower limits than the present commercial

ranges if reproducible properties are to

be insured Fortunately, the shapes of the curves in Figs 6 and 7 are fairly flat, and thus the composition specifica-tions may not need to be so narrow as

to preclude commercial control

The reasons for the effects of heat

treatment, silicon content, and rolling procedure on properties are not clear at this time In this respect the present work has raised more questions than it has answered More thorough study of some of these problems should be very rewarding More potent tools than conventional replica electron microscopy and X-ray diffraction will be required

to investigate these questions In ticular, the observation of the fine structure of martensite by transmission electron microscopy would be a likely approach

par-CONCLUSIONS

A study has been made of the effects

of composition and rolling practice on the strength properties of Type 18-8 stainless steels The results have shown:

1 Lowering the silicon content to the 0.01 to 0.13 weight per cent range sig-nificantly improved the toughness quali-ties of sub-zero rolled 18-8 stainless steels At these silicon levels notch-tensile ratios of 1.0 were obtained at yield strengths up to 274,000 psi

2 For the best combination of strength and toughness a suitable com-position appears to be approximately 7

to 8.5 nickel, 18 to 19 chromium, 0.1 to

1 manganese, 0.01 to 0.15 silicon, 0.04 to 0.08 carbon, and 0.01 to 0.04 nitrogen

3 Preliminary tests at —320 F dicated that notch-tensile ratios of 1.0

in-at yield strengths on the order 300,000 psi could be obtained with lower-silicon alloys

4 The toughness is to some extent dependent upon the rolling conditions Reductions of 40 per cent at —106 F gave comparable yield strengths but better toughness than 20 per cent reduc-tion at —320 F Residual austenite is not required for good toughness

5 The effect of composition on yield strength properties after cold rolling and heat treating 24 hr at 800 F could

be correlated in terms of the influence

of the alloying elements on the M^

temperature

6 The 800 F heat treatment after rolling apparently raised the yield strength of sub-zero rolled alloys when they contained some minimum amount

of austenite With no austenite, this heat treatment lowered the strength

7 Hardening by sub-zero rolling sults primarily from martensite forma-tion A significant amount of hardening also may result from the work hardening

re-of the austenite prior to transformation

to martensite

FIG 9—Tensile Properties of Specimens Rolled 40% at - 1 0 6 F and Heat Treated 2-1 hr at 800 F,

Showing Effect of Silicon Content

TABLE 4—RESULTS OF TENSION TESTS AT - 3 2 0 F

Elonga-21 6»

2 1^

Notch Tensile Strength (NTS), lisi

" A = 20% reduction at - 320 F plus 800 F for 24 hr

<> B = 40% reduction at - 106 F plus 800 F for 24 hr

' Broke at knife edge

R E F E R E N C E S

(I) N A Ziegler and P H Brace, "Hardening

of Austenitic Stainless Steels by

Mechani-cal Working at Sub-zero Temperatures,"

Proceedings, Am Soc Testing Mats., Vol

50, 1950, p 861

V N Krivobok and A M Talbot, "Effect

of Temperature on the Mechanical

Prop-erties, Characteristics, and Processing of

Austenitic Stainless Steel," Proceedings,

Am Soc Testing Mats., Vol 50, 1950), p

895

G B Espey, A J Rapko, and W F Brown

"Effect of Cold Rolling and Stress Relief

on the Sharp Edge Notch Tensile

Char-acteristics of Austenitic Stainless Steel

(2)

(3)

Sheet Alloys," Proceedings, Am Soc

Test-ing Mats., Vol 59, 1959, p 816

(4) G B Espey, M H Jones, and W F

Brown, "The Sharp Edge Xotch Tensile Strength of Several High Strength Steel

Sheet Alloys," Proceedings, Am Soc

Test-ing Mats., Vol 59, 1959, p 837

(5) J F Watson and J L Christian, "Low Temperature Properties of Cold Rolled AISI T>T3es 301, 304 ELC, and 310 Stain-

less Sheet Steel," ASTM STP No 287,

Am Soc Testing Mats., 1960, p 170

(6) C R Mayne, "Effect of Zerolling on the Properties of Modified Type 347 Stainless

Steel," ASTM STP Xo 287, Am Soc

Testing Mats., 1960, p 150

(7) D W McDowell and J R Mihalisin,

"Zerolling Boosts Strength of Stainless

Steel," Materials in Design F.ngineering,

Vol 53, May, 1961, p 10

(8) M P Hansen, "Smooth and Sharp Notch Tensile Properties of Cold Reduced AISI

301 and 304L Stainless Steel at 75, - 3 2 0 ,

and - 4 2 3 ° F , " NASA Technical Note 5P2, 1961

D-(9) G B Espey, M H Jones, and W F Brown, "Some Factors Influencing the Fracture Toughness of Sheet Alloys for Use in Lightweight Cryogenic Tankage,"

ASTM STP No 302, Am Soc Testing

Mats., 1961

(10) J F Watson and J L Christian, "A Study of Austenite Decomposition at

Cryogenic Temperatures," Transactions,

Met Soc of AIME, Vol 224, 1962, p 998 (11) K M Carlsen and K C Thomas, "Effect

of Composition, Heat Treatment and Cold Rolling of Mechanical Properties of Cr-Ni

Stainless Steels," Transactions Quarterly,

Am Soc Metals, Vol 55, 1962, p 462 (12) "Fracture Testing of High-Strength Sheet

Materials," ASTM Bulletin, January, 1960,

p 29

(13) R P Reed, "The Spontaneous Martensitic Transformation in 18% Cr, 8% Ni Steels,"

Acta Melallurgica, Vol 10, 1962, p 865

(14) B Cina, "Effect of Cold Work on the

y-a Transformation in Some Fe-Ni-Cr Alloys," Journal of the Iron Steel Inst.,

Acta Metalhirgica, Vol 10, 1962, p 1077

(17) V F Zackay, W M Justusson, and J D Schmatz, "Deformation of Metastable Austenite—An Interim Progress Report,"

Metals Progress, Vol 80, September, 1961,

p 68

(18) W M Justusson and D J Schmatz, "Some Observations on the Strength of Martensite Formed from Cold-Worked Austenite,"

Transactions Quarterly, Am Soc Metals,

Vol 55, 1962, p 640

(19) T Angel, "Formation of Martensite in

Metastable Austenite," Journal of the Iron Steel Inst., Vol 177, 1954, p 165

(20) H C Fielder, B L Averbach, and M Cohen "The Effect of Deformation on the Martensitic Transformation in Austenitic

Stainless Steels," Transactions, Am Soc

(23) N N Lyulicheva and N V Pisareva,

"Corrosion Resistance of Austenitic Steels Worked at Sub-zero Temperatures,"

Izvest Chetn Met., February, 1960, p 78

STP369-EB/Apr 1965

THE MECHANISMS OF DEFORMATION AND WORK HARDENING

IN AISI TYPE 301 STAINLESS STEEL

BY W F BARCLAY'

SYNOPSIS The mechanical properties of austenitic and martensitic stainless steels

have been the subject of numerous research papers The most complex types of

materials, however, are the semiaustenitic steels which transform from

aus-tenite to martensite during plastic deformation The most common commercial

material which makes use of this characteristic is AISI Type 301 The research

reported here had as its goal the determination of the mechanisms which

con-trol strain hardening and plastic deformation in commercially produced Type

301 The primary research tools used to reveal the mechanisms of deformation

were the measurement of martensite forming during plastic deformation and

the examination, by electron transmission and electron diffraction, of tension

specimens deformed to various increments of plastic strain

The work of Angel (i)^ and Powell et al

(2) revealed that the transformation

from austenite to martensite during

plastic deformation varied with the mode

of deformation and was a maximum for

uniaxial tension It is usually a good

approximation to describe plastic fllow

n = work-hardening coefficient, and

a = material constant evaluated from

test data

The rate of work hardening is then

described by w and determined as the

slope of a In 5 versus In e curve

Angel directly related a change in n

with strain in 18Cr-8Ni alloys to the

fraction of martensite formed The alloy

stability could be predicted successfully

by consideration of chemical composition

as indicated by Angel (i), and Eichelman

and Hull (3) The less stable the austenite

at room temperature, the quicker it

transformed with strain, and the higher

the n value

Early research, as reported in the

literature, considered only three

mecha-nisms of deformation to be operative in

semiaustenitic stainless steels: (1)

dis-location motion (slip) in austenite, (2)

' Supervisor, H i g h - T e m p e r a t u r e Alloy a n d

Stainless Steel Research, Republic Steel Corp.,

R e s e a r c h C e n t e r , Metallurgical Div., Cleveland,

Ohio

^ T h e boldface n u m b e r s in parentheses refer

t o t h e list of references a p p e n d e d t o t h i s paper

martensite formation, and (3) tion motion (slip) in martensite This was later modified by Cina (4) who added the formation of a he.xagonal close-packed phase which accompanied martensite formation but which dis-appeared with complete transformation

disloca-to martensite This hexagonal packed phase has more recently been reported by Dash and Otte (S) in a pure 18Cr-12Ni alloy where transformation occurred by cooling below the martensite start temperature Other recent research has also revealed the structure of stack-ing faults (6) and of martensite (7,8) as revealed by electron transmission micros-copy Although martensite formation due to deformation has been studied bv electron transmission microscopy of unstable iron nickel alloys,' there has been no correlation of observed work hardening during tensile straining with structure on a fine scale as was done in the research reported here

close-E X P close-E close-E I M close-E X T A L P close-E O C close-E D U R close-E S

All testing of mechanical properties was performed at room temperature on specimens of recommended ASTAI di-mensions made from 0.020-in thick strip products produced commercially

Mechanical properties reported here were determined on an Instron tension testing machine at a constant crosshead rate of 0.05 in./min The true strain was obtained b}' measuring the minimum cross section dimensions every 5 per cent longitudinal strain as measured bv a

men A tension specimen of Type 301 was

used for a standard of 100 per cent martensite, after it had been destabilized and cold rolled to 100 per cent martensite

as indicated by X-ray measurements This specimen was used to establish a full-scale deflection range of 10 mv on a 0.1-sec full-scale response recorder which measured induced magnetization (mar-tensite content) as a function of time, The correct strain intervals were marked

on the curve manuall}' The effect ol changing specimen geometry during tension tests was taken into considera-tion by measuring the full-scale deflec-tion of various geometry 100 per cent martensite tension specimens which simulated the geometry of strained test specimens In this manner, corrections were made which allowed calculation oi the per cent martensite by volume foi dynamically strained tension specimens, This method of martensite determina-tion, as checked on identical specimens

by X-ray measurements and the volume change within the gage length, was found

to agree within ± 5 per cent

All studies of the microstructures were done on an RCA EMU 3-100 kv electron microscope Specimens were

prepared from within the gage length oi

bulk tension specimens The foil tion was performed in two stages First, chemical thinning was achieved using a

prepara-30 per cent HNO3, 15 per cent HCl, IC per cent HE, 45 per cent H2O solution

as recommended by Keown and ing (a) This reduced the metal thickness

Picker-BARCLAY ON DEFORMATION AND WORK HARDENING 27 TABLE 1—CHEMICAL COMPOSITION OF ALLOYS, WEIGHT PER CENT

Offset, psi

38 600

40 500

Tensile Strength, psi

o

0.01 0.05 0.10

FIG 1—True Stress-True Strain Plot Showing the Relation of Alloy Stability to Strain

Harden-ing and Martensite Formation

from 0.020 in to 0.003 to 0.006 in

Electropolishing followed using the

"window method" (12) with an

elec-trolyte mixture of one part perchloric

acid to 20 parts glacial acetic acid

EXPERIMENTAL RESULTS AND

DISCUSSION

The chemical composition and

me-chanical properties of the two analyses

investigated are given in Tables 1 and 2,

respectively Note that the alloy

com-positions are very similar with the

excep-tion of nickel content The alloys

repre-sent a low- and high-stability commercial

Type 301 as controlled by nickel

con-tent The log true stress-log true strain

curves are plotted in Fig 1 for both

alloys Note that the less stable alloy A

has the higher rate of work hardening as

well as the greater amount of martensite

formed with strain as would be expected

The stability of the alloys is also

re-flected by the strain at which a change in

slope occurs which coincides with the

onset of martensite transformation In a

broader range of chemistries than those

reported here, it was found that the onset

the dislocation density increased until,

at about 0.08 to 0.10 strain, dislocation tangles were forming into fairly dense cells The appearance of cells' coincided with the appearance of stacking faults as shown in Fig 3 These structures are classified as stacking faults, for they did not give twin diffraction patterns when studied by selected area electron diffrac-tion Further straining results in an in-crease in stacking fault density By the

FIG 2—Low Dislocation Densit}' in an tenite Grain of Annealed Type 301 Stainless Steel (X 66,000)

.\us-of instabihty occurred between a true strain of 0.18 and 0.30 and accompanied the gross transformation to martensite

The structure of the alloys is clearly defined by the series of electron transmis-sion micrographs shown in Figs 2-8

The annealed initial structure of very low dislocation density is seen in Fig 2

Annealing twins were often present but are not seen here With plastic strain,

FIG 3—Dislocation Tangles, Cell Forrnation, Stacking Faults, and Deformation Twins in Austenite Strained 0.10 (X66,000)

FIG 4—Heavy Dislocation Density with an Increased Amount of Twinning, 0.20 Strain (X39,250)

time a strain of 0.20 has been reached, the dislocation density in the austenite

is becoming very heavy, and tion twins of lenticular shape have formed (Fig 4) The twins have been differentiated from the stacking faults

deforma-by the selected area diffraction patterns which showed a twin orientation super-imposed on that of the face-centered cubic matrix grain The twins could be identified in electron transmission by their lenticular shape, greater thickness, and more uniform contrast density than

28 ADVANCES IN THE TECHNOLOGY or STAINLESS STEELS

FIG 5—An Example of Stacking Faults and

Deformation Twins Formed During 0.30 Strain

(X 65,450)

slightly less than that in the surrounding austenite The increase in strain from 0.30 to 0.60, represented by Figs 5, 6, and 7, resulted in a greater quantity of martensite, more twinned austenite, and

a dislocation density in the austenite which was so heavy that it prevented detailed study In this region of increased strain, an increase in the dislocation density of the martensite was apparent

Strain above 0.60 resulted in an tremely heavy dislocation density ac-companied by a twin-like structure in the martensite This is best illustrated

ex-by Fig 8 The twinning is on a very fine scale and cannot be positively identified

by electron diffraction The twin-like structure had an appearance similar to that of twinned martensite observed in

FIG 6—Deformation Martensite in a Matri.x

the stacking faults It was also between

0.20 and 0.30 strain that the first

mar-tensite plates were observed These

plates contained a very heavy

disloca-tion density which appeared to be

NOTE—Sample taken near the fracture area

at 0.72 strain

FIG 8—An Extremely Heav\- Dislocation Density in Deformation Martensite Which Also E.xhibits Twinned Martensite (X 65,450)

iron-30.9 Ni alloys by Warlimont (13) and Richman (14) The twins were, however, much smaller than those re-ported by either author The twins in the martensite may be due either to the high dislocation density in the austenite before martensite formation, or they may occur with strain subsequent to mar-tensite formation

CONCLUSIONS The microstructure of Type 301, as it hanges with plastic deformation, is both 'complex and unique in the types of phases which appear The change of work hardening coefficient with strain has been definitely related to the forma-tion of deformation martensite A less stable alloy undergoes transformation sooner, has more martensite formation

at a given strain, and reaches a higher tensile strength and more uniform elon-gation than a more stable alloy There is

of course, an optimum austenite stability below which alloys transform to martens-ite too quickly with strain, and have a comparable tensile strength with lower elongations

The deformation of Type 301 has been observed to occur by at least six mechanisms: (1) dislocation motion in austenite; (2) dislocation tangles, cell formation, and formation of stacking faults; (3) deformation twins in austen-ite; (4) martensite formation; (5) dis-location motion in martensite; and (6) deformation twins in martensite The mechanisms are listed in order of appear-

^ c e with increasing strain It must be remembered that several mechanisms are operative simultaneously

The structures observed indicate that one volume of austenite may be expected

to proceed through Stages 1 to 6, secutively, with increasing strain The structures suggest that most of the plastic deformation occurs via deforma-tion mechanisms in the austenite Stages

con-1 through 3, plus martensite formatior

in Stage 4 In unbalanced alloys which are less stable than Type 301, more martensite forms at low plastic strains and the total strain to fracture may be 0.40 as compared to 0.68 to 0.80 for com mercial Type 301 Lean, 7.0 Ni, Type

301 analyses are observed to have ex tremely good elongation since they hav(

an optimum combination of strain in th( austenite, with complete transformatioi

to martensite before fracture and defor mation in the martensite If the alio] composition is too rich (high Ni content) the increased alloy stability results ii excessive dislocation generation an( twinning in the austenite, causing frac ture before complete transformation ti martensite can occur

REFERENCES (1) T Angel, "Martensite in Austenitic Stair

less Steels," Journal, Iron and Steel Inst

Type Stainless Steel," Transactions, An

Soc Metals, Vol 45, 1953, pp 77-95 (4) B Cina, "Effect of Cold Work on tl

7 —> a Transformation in Some Fe-Ni-(

Alloys," Journal, Iron and Steel Insi

Vol 177, No 1, 1954, pp 406-422 (5) J Dash and H Otte, "Thin Film Tran mission Observations of Athermal Tran formations in Stainless Steel," 5th Inte

BARCLAY ON DEFORMATION AND W O R K HARDENING 29

Vol I, Academic Press, New York and

London, 1962, p HH-3

(6) J A Venables, "The Observation of

Mechanical Twins in Face-Centered Cubic

Materia]," European Regional Conference

on Electron Microscopy, Delft, 1961, pp

443-446

(7) J F Breedis, "On the Nucleation of

Martensite in Thin Foils of Fe-16 Wt.%

Cr-12 Wt.% Ni," Electron Microscopy,

5th International Congress for Electron

Microscopy, Vol 1, Academic Press, New

York and London, 1962, p HH-5

(8) P M Kelly, "The Formation of

Marten-site and Bainite in Steels," Electron

Mi-croscopy, 5th International Congress for

Electron Microscopy, Vol 1, Academic

Press, New York and London, 1962, p

(10) G W Form and W M Baldwin, Jr.,

"The Influence of Strain Rate and perature on the Ductility of Austenitic

Tem-Stainless Steels," Transaction, Am Soc

Metals, Vol 48, 1956, pp 474-485

(11) S R Keown and F B Pickering, "A Chemical Technique for the Preparation

of Thin Metal Foils for Transmission

Electron Microscopy," Journal, Iron and

Steel Inst., Vol 200,' 1962, p 757

(12) H, M Tomlinson, "An Electropohshing Technique for the Preparation of Metal Specimens for Transmission Electron

Microscopy," Philosophical Magazine, Vol

3, 1958, p 367

(13) H Warlimont, "On the Martensitic

Struc-ture of Iron-30.9% Nickel Alloy," Electron Microscopy, 5 th International Congress

for Electron Microscopy, Vol 1, Academic Press, New York and London, 1962, p HH-6

(14) R H Richman, "Plastic Deformation

Modes in Fe-Ni-C Martensite," actions, Am Institute of Mining and

Trans-Metallurgical Engrs., Vol 227, 1963, pp 159-166

ously in a controUed-atmosphere furnace (depicted) The furnace atmosphere

system includes an ammonia storage tank, a dissociator, and an adsorption

dryer Furnace capacity depends upon its specific size, source, temperature,

and cooling rate, as well as the emissivity of the material processed Atmosphere

problems are purity and nitrogen pickup Purity of the dissociated ammonia

gas depends upon initial raw ammonia contaminants and dissociator

perform-ance Nitrogen pickup can be prevented by annealing in purified hydrogen

atmospheres Production problems include atmosphere system leakage,

main-tenance of a temperature head, strip cleaning problems, and crystallized

sur-face Fabrication problems have included soldering and drawing The

advan-tages of bright annealed strip are improved surface finish, reduced buffing

costs, and improved corrosion resistance

Bright annealing, as referred to in this

paper, is the annealing of stainless steel

sheet and strip continuously in a

con-trolled-atmosphere furnace This

pro-tective furnace atmosphere protects the

steel from oxidation, thereby

elimi-nating the usual pickling operations

associated with conventional annealing

The surface finish, after bright annealing,

is the same as it was in the "as cold

rolled" condition The furnace

atmos-pheres are usually dry hydrogen or dry

dissociated ammonia which is composed

of 75 per cent hydrogen and 25 per cent

nitrogen

To produce the stainless steel

re-quired by the automotive industry, it

was necessary to do considerable

polish-ing and buffpolish-ing to impart a satisfactory

corrosion resistance, but competitive

and economic conditions in the industry

demanded corrosion resistance with

mini-mum buffing Various investigators found

that conventionally annealed and bright

pickled Type 430 stainless steel resulted

in a chromium-depleted surface with

reduced corrosion resistance In addition,

it was found that annealing in an inert

atmosphere, such as dry dissociated

ammonia or hydrogen, did not result

in a chrome-depleted surface

The joint efforts of the stainless

producers and the builders of industrial

in regard to operating costs, to meeting annealed properties, and so on, and yet would not change the surface, thus solving the corrosion (chrome-depletion) problem

B R I G H T ANNEALING U N I T The bright annealing unit is composed

of the following major components (Fig

1):

A Payoff reel or uncoiler

D Shear—for preparation of the ends

of the coils for welding

K Welding unit—for welding of the

coils for continuous operation

F Strip or sheet cleaning unit—usually

alkaline cleaner or trichlorethylene

G Accumulators:

Entrance accumulator—main pose to allow for time to weld the coils together

pur-Exit ,accum,ulator—main purpose

to allow time for the removal of the coil from the take-up reel

3 Annealing zone—containing inert

atmosphere of dry dissociated ammonia

or hydrogen M a y be of muffle type or high purity refractory brick lined furnace type

4 Cooling zone—containing inert

at-mosphere of dry dissociated ammonia

to hydrogen and nitrogen, and an ing dryer of the molecular sieve type with fully automatic controls for the further reduction of the dew point of the dissociated ammonia

adsorb-The ammonia dissociator consists essentially of one or more retorts filled with a catalyst, which is heated exter- nally and through which gaseous am- monia is passed The ammonia disso- ciates according to the reaction 2NH3 - ^ 3H2 + N 2 For any given conditions, the per cent of dissociation will depend upon the temperature, the rate of flow, and the condition of the catalyst Brief descriptions of various parts of the dissociator follow:

1 Pressure Reducing Valve—The

am-monia vapor supply enters the ciator through a constant-pressure reduc- ing valve This reducing valve is set for the minimum inlet pressure to the retorts (not to exceed 30 psi max) necessary to obtain the rated capacity of the dissociator with a back pressure of 5 psi a t the throttling valve A pressure relief valve located downstream from the pressure reducing valve will open

disso-if the pressure at this point exceeds the maximum of 30 psi A pressure gage indicates the pressure a t the above point

2 Dissociator Retort—After passing

through the pressure reducing valve, the ammonia enters the dissociator retorts which are filled with a catalyst The dissociator retorts are located in a refractory-lined heating chamber, and heat is supplied for the dissociating process by means of nozzle mix burners which operate on "high-low" control From the dissociator retorts, the disso-

BOYER AND P E R R I N E ON BRIGHT ANNEALING OF STAINLESS S T E E L 31 dated ammonia goes to a water-cooled

surface cooler and then passes through

the flow meter and into the service line

3 Safely Alarm System—In order for

the ammonia dissociator to operate

efficiently, the temperature of the

disso-ciator retorts must be at the proper

value Too low an operating temperature

will leave a high percentage of residual

ammonia in the dissociator; on the other

hand, too high an operating temperature

will decrease the life of the alloy retorts

and the catalyst Temperature control

pends upon the specific size, source temperature, and equally important, its ability to cool the work below a certain minimum temperature to prevent oxida-tion Another factor which has a signif-icant bearing on the production capacity

is brightness This inherent factor, known

as absorptivity or emissivity, varies from 0.12 for mirror bright to 0.3 for dull stainless steels and may sometimes

be outside of this range

Annealing to obtain certain physical properties requires that the steel be

mathematical expression derived from the following equations:

J - COILING REEL K- WELDER

ENTRY END

® © ®

®®^A^^

FIG 1—Schematic Arrangement of Bright Annealing Line

equipment is provided to hold the

tem-perature of the dissociator at the desired

Dperating value The alarm system will

Dperate if the temperature of the

disso-:iator varies more than the permissible

preset values An ammonia pressure

^age with high- and low-pressure

con-tacts is provided with related alarm

system An air-operated alarm horn is

arovided in case the burner gas pressure

"ails or is shut off

OPERATIONAL AND CONTROLLING

FACTORS

Furnace Performance:

The capacity of a given furnace

de-heated above the recrystallization perature or slightly higher, followed by a short soak and cool, or a cool imme-diately To obtain maximum production, the furnace is operated at the maximum strip speed which will permit the strip

tem-to reach the desired temperature Actual time to reach temperature sets the strip speed, and this time depends upon the source temperature Furthermore, the time is directly proportional to the den-sity, specific heat, and thickness of the strip, and is inversely proportional to the emissivity

The time to heat a given strip to some temperature can be expressed by a

ings, and

Ti = absolute temperature of work

By working with a square foot of metal 1 in thick and expressing time in

seconds, we can obtain new values for q and Q represented by q^ and Q^ respec-

tively